WO2021232962A1 - Target site prediction method and apparatus, and electronic device and storage medium - Google Patents

Abstract

The present application relates to the technical field of electronic devices, and provides a target site prediction method and apparatus, and an electronic device and a storage medium. The target site prediction method comprises: obtaining current arrival information, the current arrival information comprising current arrival time and a current arrival point; obtaining, according to a historical riding record, a target historical arrival point matching the current arrival time; when the current arrival point is the same as the target historical arrival point, predicting a target site according to the historical riding record, the current arrival time and the current arrival point; and when the current arrival point is different from the target historical arrival point, predicting a target site according to the historical riding record and the current arrival point. The method can improve the prediction accuracy of the destination site.

Description

Target site prediction method, device, electronic equipment and storage medium

Cross-references to related applications

This application claims the priority of the Chinese application with application number 202010437224.4 filed on May 21, 2020, which is hereby incorporated by reference in its entirety for all purposes.

Technical field

This application relates to the technical field of electronic equipment, and more specifically, to a method, device, electronic equipment, and storage medium for predicting a target site.

Background technique

At present, the functions of mobile terminals are becoming more and more abundant. In order to be able to use mobile terminals to better serve users, it is often necessary to predict the destination point the user will reach. However, in the prediction method of the relevant destination point, there is still a problem of insufficient accuracy.

Summary of the invention

In view of the above-mentioned problems, this application proposes a method, device, electronic equipment, and storage medium for predicting a target site, which can improve the above-mentioned problems.

In the first aspect, an embodiment of the present application provides a method for predicting a destination station. The method includes: acquiring current station entry information, where the current station station information includes the current station station time and the current station station; and according to historical ride records, Obtain the target historical entry point matching the current pit stop time; when the current entry point is the same as the target historical entry point, according to the historical ride record, the current pit stop time and the current entry point Predict the destination station; when the current inbound station is different from the target historical inbound station, predict the destination station based on the historical ride records and the current inbound station.

In the second aspect, an embodiment of the present application provides a device for predicting a destination site. The device includes: a current inbound acquisition module, configured to obtain current inbound information, where the current inbound information includes the current inbound time and the current inbound time. Inbound; the target inbound acquisition module is used to obtain the target historical inbound station that matches the current inbound time according to the historical ride records; the first prediction module is used to when the current inbound station and the target historical inbound When the stations are the same, predict the destination station based on the historical ride record, the current station time and the current station; the second prediction module is used when the current station and the target historical station are not the same , Predict the destination station based on the historical ride records and the current inbound station.

In a third aspect, an embodiment of the present application provides an electronic device, including: one or more processors; a memory; one or more application programs, wherein the one or more application programs are stored in the memory and It is configured to be executed by the one or more processors, and the one or more application programs are configured to execute the method for predicting the target site provided in the first aspect described above.

In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium. The computer-readable storage medium stores program code, and the program code can be invoked by a processor to perform the purpose provided in the first aspect. The prediction method of the site.

Description of the drawings

In order to more clearly describe the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings that need to be used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

Fig. 1 shows a flowchart of a method for predicting a destination site according to an embodiment of the present application.

Figure 2 shows a schematic diagram of a database cloud architecture according to the present application.

Fig. 3 shows a flow chart of a method for predicting a destination site according to another embodiment of the present application.

Fig. 4 shows a flowchart of step S220 in the method for predicting a destination site according to another embodiment of the present application.

Fig. 5 shows another flow chart of step S240 in the method for predicting a destination site according to another embodiment of the present application.

Fig. 6 shows a flowchart of a method for predicting a destination site according to another embodiment of the present application.

Fig. 7 shows another flow chart of step S330 in the method for predicting a destination site according to another embodiment of the present application.

Fig. 8 shows a flow chart of a method for predicting a destination site according to still another embodiment of the present application.

Fig. 9 shows a schematic diagram of an application scenario according to the present application.

Fig. 10 shows a flowchart of a method for predicting a destination site according to still another embodiment of the present application.

Fig. 11 shows a schematic diagram of the coverage area of a base station according to the present application.

Fig. 12 shows a schematic diagram of the overall flow of a base station positioning method according to the present application.

FIG. 13 shows a schematic block diagram of the overall flow of a method for predicting a destination site according to an embodiment of the present application.

Fig. 14 shows a block diagram of a device for predicting a destination site according to an embodiment of the present application.

FIG. 15 is a block diagram of an electronic device for executing the method for predicting a destination site according to an embodiment of the present application according to an embodiment of the present application.

Fig. 16 is a storage unit for storing or carrying program codes for realizing the prediction method of the destination site according to the embodiment of the present application according to an embodiment of the present application.

Detailed ways

In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the accompanying drawings in the embodiments of the present application.

At present, the destination prediction method is usually based on a fixed device or other specific applications to obtain user data for prediction and estimation. It involves sensitive information such as real-time location tracking of the user. In addition, the user needs to provide clear departure and destination information to be effective. The implementation of services for specific user groups.

In addition, the inventor has discovered through long-term research that the existing destination prediction method is usually to make simple statistics on the user's historical destinations, and sort the use times of the user's historical destinations in a certain time in descending order, and select The most frequently used historical destination is used as the predicted travel destination; or the last use time of the user's historical destination is sorted from nearest to farthest, and the most recently used historical destination is selected as the predicted travel destination. The existing destination prediction methods do not consider the impact of the user's travel time and the user's place of departure on the destination prediction, resulting in inaccurate prediction results, and there is no technical barrier or leading edge.

In addition, the data acquisition in existing prediction methods includes base station, WiFi (Wireless Fidelity), Bluetooth, GPS (Global Positioning System, Global Positioning System) location data, time data, and vehicle id (Identity document, unique identification) Number) and so on, more dependent on user interaction. In addition, GPS cannot cover all corners of the environment, and its accuracy is limited. WiFi and Bluetooth information rely on external equipment, which is costly and unfriendly to the outdoor environment. In addition, the existing destination prediction schemes usually only target users such as cars, and few or fewer provide corresponding services for users who take subways.

Therefore, after long-term research, the inventor has discovered and proposed the method, device, electronic equipment, and storage medium for the prediction of the destination site provided in the embodiments of this application. If the time forecast is not very accurate, you can directly forecast based on the current entry site. Improve the accuracy of the destination site prediction, making the predicted destination more in line with the actual needs of users. In addition, the prediction method of the destination site in this application is also applicable to subway scenes. The specific target site prediction method will be described in detail in the subsequent embodiments.

Please refer to FIG. 1, which shows a schematic flowchart of a method for predicting a destination site according to an embodiment of the present application. The prediction method can be applied to electronic equipment. The electronic device can be a mobile terminal, so that the prediction method can be executed locally; the electronic device can also be a server, so that the prediction method can be executed in the cloud. In a specific embodiment, the prediction method of the destination site can be applied to the prediction device 700 of the destination site as shown in FIG. 14 and the electronic equipment configured with the prediction device 700 of the destination site (FIG. 15 ). The following will elaborate on the process shown in Figure 1. The prediction method of the destination site shown may specifically include the following steps:

Step S110: Obtain current inbound information, where the current inbound information includes the current inbound time and the current inbound station.

In the embodiment of the present application, the electronic device may obtain the user's current inbound information, so as to predict the destination site based on the current inbound information. Among them, the current inbound information may include the current inbound time and the current inbound station.

In some embodiments, the user's current inbound information can be obtained in multiple ways. For example, the user’s current entry information can be identified by swiping identification, NFC (Near Field Communication) identification, GPS identification of the inbound position, and subway driving identification.

In some embodiments, a preset application may be installed on the mobile terminal. When the user enters the preset traffic scene by swiping the code with the mobile terminal, the preset application can identify the name, route, city, and time of the station entered. . Among them, the preset application can be configured to be suitable for running in a preset traffic scene, and the preset traffic scene can be any traffic scene such as subway, bus, high-speed rail, train, and airplane.

In some implementation manners, the preset application may be an application program such as a ride application, a payment application, etc., which is not limited in the embodiment of the present application.

Exemplarily, the preset application may be a bus ride application, which is suitable for running in a bus scene. Among them, the bus ride application can be used to generate a ride logo for users to swipe the code to ride the bus. When a user uses a mobile terminal to swipe a code to board a bus, the name of the site, route, city, and time of the user's boarding can be identified through the bus ride application.

Exemplarily, the preset application can also be a payment application. The payment application can be configured with multiple e-cards. The e-card can be a radio frequency card simulated based on a traffic card and can support NFC payment. In some examples, the mobile terminal can be built-in The NFC module can interact with the NFC reading device based on the NFC module, so that when the user takes a transportation, the user can move the mobile terminal close to the NFC reading device, and the NFC reading device can read the configuration of the payment application The transportation card to make the payment. The mobile terminal can obtain the inbound information according to the payment status of the payment application.

As a specific implementation manner, an inbound identification module may be configured on the mobile terminal, and the inbound identification module may be embedded in the android system. In the subway traffic scenario, when a user uses an APP (Application) installed on a mobile terminal to enter a subway station by swiping the code, for example, the Shanghai Metro is a metropolis, Shenzhen has a WeChat applet, Shenzhen Metro, and Beijing uses Yitongxing. After the user swipes the code to enter the station, the mobile terminal will pop up the successful entry, the entry time and the entry site on the mobile terminal. The inbound recognition module configured on the mobile terminal can recognize the changes on the screen of the mobile phone to obtain the current inbound information including the inbound time and the inbound station.

In some embodiments, after the current inbound information is obtained, the current inbound information can be stored and recorded, so as to continuously update the database content storing historical ride records, so that more accurate updates can be made based on the continuously updated database content. The prediction of the destination site in line with the current scenario. As a way, the acquired current inbound information can be recorded in the format InStation={time,name}, where InStation is the current inbound information, time is the current inbound time, and name is the current inbound station.

Step S120: According to the historical ride records, obtain the target historical entry point matching the current entry time.

In some scenarios, when entering the same inbound site, the user's inbound time may be unstable and fluctuate for a period of time, and sometimes may not be within the long-term inbound time that users are accustomed to. In this case, if combined Predicting the destination site at the current stop time may cause the predicted destination site to be inaccurate and deviate from the actual needs of users.

Therefore, in this embodiment of the application, after obtaining the current inbound information, the current inbound time can be first predicted according to the current inbound time, so as to determine whether to perform the destination site based on the current inbound time according to the predicted result of the inbound station. predict. Specifically, according to the historical ride records, the target historical entry point matching the current entry time can be obtained. Among them, the target historical entry site can be understood as the predicted entry site after long-term independent learning of historical ride records.

In some embodiments, it is possible to determine the target historical entry point that matches the current entry time based on multiple historical entry times and multiple historical entry points in the long-term historical ride record. As a way, you can obtain the target historical inbound time that matches the current inbound time from multiple historical inbound times, and then use the historical inbound time corresponding to the target historical inbound time as the current inbound time The matched target historical entry site, that is, the predicted current entry site. Among them, the target historical inbound time that matches the current inbound time may be the same historical inbound time as the current inbound time, or it may be the closest historical inbound time to the current inbound time.

For example, the current pit-stop time is 8:45 am, the target historical pit-stop time that matches the current pit-stop time can be the closest 9 am from 8:45 am, and then the historical date can be 9 am The ride record of the station is extracted, and the entry site entered at 9 am in the ride record is used as the target historical entry site, that is, the predicted current entry site.

Further, since there may be different historical entry points corresponding to the same time point on different historical dates, there may be multiple historical entry points corresponding to the target historical entry time. For example, the entry and exit station of the work travel route corresponding to 8 a.m. on weekdays is likely to be different from the entry station of the entertainment travel route corresponding to 8 a.m. on weekends. Therefore, in some embodiments, when there are multiple historical entry sites corresponding to the target historical entry time, the most likely historical entry site can be selected from them as the target historical entry site. As a way, it is possible to select the historical inbound stop with the highest number of occurrences from the multiple historical ride records corresponding to the target historical inbound time as the target historical inbound stop.

In some embodiments, a database may be established to record and store the outbound information and inbound information of the user from the current date on each trip in the past. Wherein, the database can be stored locally or in the cloud, which is not limited here.

Exemplarily, assuming a subway scene, as shown in FIG. 2, FIG. 2 shows a cloud storage architecture of a subway habit database. Through this storage architecture, automatic acquisition and storage of ride records can be realized. Among them, the specific functions of each module of the architecture are as follows:

Station recognition module 201: Used to recognize when users enter a subway station. When a user enters a subway station with a mobile phone, he can recognize the name, line, city, and time of entering the subway station through the application.

Outbound module 202: Used to identify the user when leaving the subway station. When the user leaves the subway station by swiping the code with the mobile phone, the name, line, city, and time of leaving the subway station can be identified through the application.

Data cloud storage module 203: used for data storage. When the station entry information is detected (the user can scan the code to enter the subway station can be detected), the corresponding information can be stored, such as the name, line, city and time of the subway station, etc. information. When detecting leaving a subway station, if the outbound information can be paired with the detected inbound information, it will be stored directly; if the inbound information and outbound information cannot be paired, they will be stored separately. Among them, it can be judged whether the inbound and outbound information matches by judging whether the brush code inbound and outbound appear in pairs, and whether the difference between the used time and the subway running time is less than a certain threshold (which can be preset). In some embodiments, the locally stored data can be uploaded to the cloud database according to the cloud collaboration strategy.

Database module 204: an information database used to store the user's subway ride history. The content of the database may include information such as location (city), event (inbound, outbound), time (UNIX, accurate to the second).

In some embodiments, the historical ride record may include inbound information and outbound information. Since there may be cases where the inbound information or the outbound information is not recognized and acquired, the historical ride records may also include only the inbound information and only the outbound information.

Illustratively, assuming a subway scene, the ride record can be stored as shown in Table 1. Among them, the historical ride records may include information such as city, inbound, outbound, and time (UNIX, accurate to the second). ‘*’ means that the information is empty, and each line represents a matching entry and exit information.

Table 1

city	Pit stop	time	Outbound	time
Shanghai	Line 2-Lujiazui	1569888000	Line 2-Songhong Road	1569889800
Shanghai	Line 2-Songhong Road	1570408200	*	*
Shanghai	*	*	Line 10-Yu Garden	1570325400

In some embodiments, when the destination site is predicted, historical ride records with empty information in the database can be eliminated, which can ensure the effectiveness of the prediction.

Step S130: Determine whether the current incoming site and the target historical incoming site are the same. If yes, step S140 can be executed; if not, step S150 can be executed.

Step S140: Predict the destination station based on the historical ride record, the current station arrival time, and the current station station.

Step S150: Predict the destination station based on the historical ride records and the current inbound station.

In the embodiment of the present application, after obtaining the target historical inbound site matching the current inbound time, it can be determined whether the current inbound site and the target historical inbound site in the acquired current inbound information are the same. And when the current station is the same as the target historical station, the destination station can be predicted based on the historical ride records, the current station time and the current station. When the current station is different from the target historical station, the destination station can be predicted based on the historical ride records and the current station. The current station time does not participate in the prediction of the destination station.

It is understandable that if the current inbound station (that is, the target historical inbound station) predicted based on the current inbound time is different from the actual current inbound station (that is, the current inbound station in the current inbound information obtained above), It can indicate that the current pit-stop time may not be the pit-stop time that the user is accustomed to for a long time at the current station, and the current pit-stop time is not suitable for participating in the prediction of the destination station based on habit. Forcible participation may lead to inaccurate prediction results and deviate from the actual needs of users. . Therefore, when the current incoming site is different from the target historical incoming site, the destination site can be directly predicted based on the current incoming site, which effectively guarantees the accuracy of the prediction result.

If the current inbound time predicted based on the current inbound time is the same as the actual current inbound station, it can indicate that the current inbound time may be the user’s long-term habit of entering the site in the current period, and the current inbound time is suitable for participation based on The prediction of the habitual destination site. Therefore, when the current station is the same as the target historical station, the destination station can be predicted based on the historical ride records, the current station time and the current station, which effectively guarantees the accuracy of the prediction result.

In some embodiments, predicting the destination station based on historical ride records, current pit stop time, and current pit stop may be to obtain one or more ride records matching the current pit stop time and current pit stop from the historical ride record, according to The outbound information in the one or more ride records predicts the target station. As a way, it may be to select the outbound station with the highest occurrence from the outbound information in one or more ride records as the predicted target station. The specific method of predicting the destination station based on historical ride records, current arrival time, and current arrival station may not be limited in this application. For example, a habit-based Markov prediction model for the target site can be established, and the Markov method can be used to predict the time point of entering and leaving the station and the probability of entering and leaving the station.

In some embodiments, predicting the destination station based on the historical ride records and the current station entry may be to obtain one or more ride records matching the current station entry from the historical ride records, and according to the outbound station in the one or more ride records Station information, predict the target station. As a way, it may be to select the outbound station with the highest occurrence from the outbound information in one or more ride records as the predicted target station. The specific method of predicting the destination station based on the historical ride records and the current station entry is not limited in this application. For example, a habit-based Markov prediction model for the target site can be established, and the probability of entering and leaving the site can be predicted by applying the Markov method.

The method for predicting the destination station provided by the embodiment of this application obtains the current station information, which can include the current station time and the current station, and then obtains the target that matches the current station time according to the historical ride records Historical entry site, to predict the destination site based on historical ride records, current entry time, and current entry site when the current entry site is the same as the target historical entry site, and when the current entry site is different from the target historical entry site , The destination station can be predicted based on the historical ride records and the current station. In this way, taking into account the influence of the current pit stop time on the prediction result, if the forecast of the entry station based on the current pit stop time is accurate, the forecast can be based on the current pit stop time and the current entry station, and based on the current pit stop time In the case that the predicted entry site is not very accurate, it can also be directly predicted based on the current entry site. Improve the accuracy of the destination site prediction, making the predicted destination more in line with the actual needs of users.

Please refer to FIG. 3, which shows a schematic flowchart of a method for predicting a destination site according to another embodiment of the present application. The following will elaborate on the process shown in FIG. 3, and the prediction method of the destination site shown may specifically include the following steps:

Step S210: Obtain current inbound information, where the current inbound information includes the current inbound time and the current inbound station.

Step S220: According to the historical ride records, obtain the target historical entry point matching the current entry time.

In the embodiment of the present application, step S210 and step S220 can refer to the content of the foregoing embodiment, and will not be repeated here.

In some embodiments, due to the randomness of the time point of entering the station, the pit time points of historical ride records may be fragmentary and continuous. For the site, the prediction results are not typical and have little reference significance. Therefore, in some embodiments, all historical inbound times in the historical ride records can be classified in time segments to obtain multiple time period categories (or time interval categories), which can be based on the time that matches the current inbound time. Segment, to predict entering the site. Since there are multiple time points in each time period, each time period can correspond to multiple historical ride records. Through multiple historical ride records rather than fragmentary historical ride records, predicting entry into the station can make the predicted results have a certain typical Sex, with reference significance.

Specifically, referring to FIG. 4, step S220 may include:

Step S221: Obtain an inbound time period matching the current inbound time according to the historical ride record, where the inbound time period is obtained by dividing all historical inbound times in the historical ride record into sections.

In some embodiments, acquiring the inbound time period that matches the current inbound time may be acquiring the time period closest to the current inbound time. For example, the current inbound time is 8:45 in the morning, and the inbound time period that matches the current inbound time can be the time period from 7:00 to 9:00. The time period may be obtained by dividing all historical inbound times in the historical ride records into sections.

As an implementation manner, all historical arrival time data may be extracted from a database storing historical ride records, and then a clustering algorithm is used for analysis to obtain multiple time period categories. For example, categorize the historical inbound time from 6:00 to 10:00 as one time period, and categorize the inbound time from 10:00 to 14:00 as another time period. Multiple time period categories can be recorded as 1, 2, 3,..., KI, and KI is the total number of time period categories. Among them, the clustering algorithm is an algorithm that involves grouping data in machine learning. A given data set can be divided into different groups through the clustering algorithm.

It is understandable that the foregoing method of time segmenting all historical inbound times is only an example, and it is not limited here. For example, it is also possible to divide 24 hours a day with a fixed time length to obtain multiple time periods. If the 24 hours a day is divided into one hour, 24 time periods can be obtained.

In some embodiments, many categories may be generated during the clustering process, but not all categories are uniformly distributed. In order to ensure the accuracy of the prediction, a category with a relatively small probability can be eliminated. As a way, the probability of each time period category can be counted first, and then only the time period category ki whose probability is greater than a specified threshold (denoted as Thr) is selected.

Optionally, the prior probability of each time period category can be counted according to the historical ride records in the data. The prior probability can be understood as the ability of each category to affect the prediction result, that is, when the prior probability is large, the greater the impact of the category on the prediction result. The calculation formula is as follows:

in,

It can be understood as the number of ride records whose historical pit stop time t is classified as k category in the database,

It can be understood that the historical pit stop time t is classified as 1, 2, 3,..., the number of KI ride records, or it can be understood as the number of all ride records in the database. Therefore, the time period category k corresponding to P(k) that is less than the specified threshold can be eliminated.

Step S222: According to the historical ride records, obtain the entry probabilities of entering different historical entry stops during the entry time period.

In some embodiments, after obtaining the inbound time period that matches the current inbound time, you can first obtain each history of the historical inbound time in the inbound time period from the database storing historical ride records. Enter the number of records in the station, and then calculate the ratio of the number of records in each historical station to the total number of historical ride records, and then the probability of entering each historical station in the inbound time period can be obtained.

Optionally, the mathematical formula can be expressed as: p(k,c), c=1, 2,...,C, where C is the total number of subway stations. p(k,c) can be understood as the probability of entering station c when the inbound time is time class k, that is, p (the inbound time class is k and inbound station=c).

In some embodiments, it is also possible to pre-calculate the inbound probability of entering different historical inbound stations in each time period, so that when the inbound time period that matches the current inbound time is obtained, the and The inbound probability of different historical sites corresponding to the inbound time period.

Step S223: Obtain the historical entry site corresponding to the entry probability with the largest value as the target historical entry site.

After obtaining the inbound time period, after entering the inbound probabilities of different historical inbound stations, the historical inbound probability corresponding to the largest numerical value can be used as the target historical inbound station, that is, the predicted inbound station. It is understandable that when the probability of entering a certain historical entry site is the greatest, it indicates that the user is likely to enter the historical entry site habitually during the subsequent entry period. Therefore, the historical entry site can be habitually entered. The entry site is the predicted entry site that the user will enter.

In some embodiments, the target historical entry site can also be directly predicted by the following calculation formula:

Among them, InStation.time is the current inbound time in the current inbound information obtained above, and Ki is the above-mentioned time period category after all historical inbound times are divided into time periods. TEStation enters the site for the predicted target history.

The current pit-stop time used to calculate the distance is the nearest time period class k, p(k, c) is used to calculate the pit-stop probability of entering station c when the pit-stop time is the nearest time class k,

It is used to obtain the site c corresponding to the maximum probability p(k,c).

It should be noted that st is the abbreviation of subject to, and its meaning is "to satisfy...", which can be understood as satisfying the second formula

Under the conditions, calculate the first formula

Step S230: Determine whether the current incoming site and the target historical incoming site are the same. If yes, step S240 to step S250 can be performed; if not, step S260 to step S290 can be performed.

Step S240: Obtain a first probability corresponding to the current inbound station according to the historical ride record, where the first probability is the number of times from the current inbound station to each historical outbound station at the current inbound time Probability.

Since the actual current inbound site is the same as the predicted target historical site inbound, it can be considered that the current inbound time is suitable for participating in the prediction of the target site based on habit. Therefore, the destination station can be predicted based on the historical ride records, the current arrival time and the current arrival station, which effectively guarantees the accuracy of the prediction result. Optionally, the current incoming station is the same as the target historical incoming station, which can be expressed by a mathematical formula: InStation==TEStation.

In some embodiments, the first probability corresponding to the current inbound station may be obtained according to the historical ride records. The first probability is the probability of going from the current inbound station to each historical outbound station at the current inbound time. As a way, you can obtain the statistics of the number of historical ride records from the current inbound station to each historical outbound station under the current inbound time, and calculate the ratio of the number of each record to the total number of historical records. You can get the first probability of going from the current inbound station to each historical outbound station under the current inbound time.

In some embodiments, the first probability corresponding to the current inbound station may be correspondingly obtained according to different historical outbound times. Specifically, referring to FIG. 5, step S240 may include:

Step S241: Obtain a historical pit stop time matching the current pit stop time according to the historical ride record.

Step S242: Obtain the transition probability from the current inbound station to each historical outbound station under the historical inbound time and different historical outbound times as the first probability corresponding to the current inbound station.

In some embodiments, Markov property and historical ride records can be used to statistically analyze the probability of going from the current inbound station to each historical outbound station at the current inbound time. Specifically, you can first obtain the historical inbound time that matches the current inbound time according to the historical ride records, and then obtain the historical inbound time and different historical outbound times from the current inbound station to each historical outbound station. Transition probability, as the first probability corresponding to the current entry site.

As a way, you can separately count the number of historical ride records from the current inbound station to each historical outbound station under the current inbound time and different historical outbound times, and calculate the number of each record and the total historical records separately The ratio of the number of entries can get the first probability of going from the current inbound station to each historical outbound station under the current inbound time and different historical outbound times.

In some embodiments, since the outbound time is also random, it may also be fragmented and continuous, which may result in the need to calculate a large number of transition probabilities. Therefore, all historical outbound time data can also be divided into time periods to obtain multiple outbound time periods. As a way, it is possible to extract all historical outbound time data from a database storing historical ride records, and then use a clustering algorithm for analysis to obtain multiple outbound time period categories. Multiple time period categories can be recorded as 1, 2, 3,..., KO category, and KO is the total number of time period categories. For the specific division method, please refer to the division of historical inbound time, which will not be repeated here.

Therefore, according to the historical ride records, the inbound time period that matches the current inbound time can be obtained, and then the transfer from the current inbound station to each historical outbound station under this inbound time period and different outbound time periods can be obtained The probability is used as the first probability corresponding to the current incoming station. As a way, it is possible to obtain the number of ride records from the current inbound station to the target historical outbound station c under the inbound time period and the target outbound time period t, and calculate the number of ride records and the total record The ratio of the number of entries can be used to obtain the transition probability from the current inbound station to the target historical outbound station c under the inbound time period and the target outbound time period t, where the value of t can be 1, 2, 3, ..., KO, c can take s=1,2,...,C, C is the total number of subway stations. In this way, multiple first probabilities corresponding to the current inbound station in different outbound time periods and different outbound stations can be obtained.

In some embodiments, the historical ride records stored in the database can also be pre-calculated in each inbound time period and each outbound time period from different historical inbound stations to different historical outbound stations. When the inbound time period and the current inbound station that match the current inbound time are obtained, the transition probability of different historical outbound stations and different historical outbound times corresponding to the inbound time period and the current inbound station can be directly obtained . Optionally, the mathematical formula can be expressed as: p(c,k,s,t), c=1,...,C,k=1,...,KI,s=1,...,C; t=1, …, KO, where c is the inbound station, k is the inbound time period category, s is the outbound station, t is the outbound time period category, and C is the total number of subway stations.

In the same way, in some embodiments, time classes with a relatively small probability may also be eliminated. As a way, the prior probability of each time period category can be counted according to the historical ride records in the data. The calculation formula is as follows:

in,

It can be understood as the number of ride records whose historical outbound time t is classified as category k in the database,

It can be understood that the historical outbound time t is classified as 1, 2, 3,..., the number of ride records in the KO category, or it can be understood as the number of all ride records in the database. Therefore, the time period category k corresponding to P(k) that is less than the specified threshold can be eliminated.

Step S250: Obtain the historical outbound site corresponding to the first probability with the largest value as the predicted destination site.

After obtaining the first probability of going from the current inbound station to each historical outbound station at the current inbound time, the historical outbound station corresponding to the first probability with the largest value can be used as the predicted destination station. It is understandable that the first probability of going from the current inbound station to a historical outbound station at the current inbound time is the greatest, which indicates that the user is likely to get used to it when he subsequently enters the current inbound station at the current inbound time. To go to the historical out-of-site site in a specific manner, therefore, the historical out-of-site site can be used as the predicted destination site that the user will reach.

In some embodiments, the target site can also be directly predicted by the following calculation formula:

Among them, EStation is the predicted destination site,

It is used to obtain the outbound station s and outbound time t corresponding to the maximum probability p(c, k, s, t).

In some embodiments, the current arrival time can also be predicted. Specifically, when the first probability of going from the current inbound station to each historical outbound station under the current inbound time is obtained, the historical outbound time corresponding to the first probability with the largest value can be used as the predicted arrival time. As a way, the inbound time period class corresponding to the current inbound station and the current inbound time can be substituted into c and k in formula (4), so as to obtain the maximum probability p(c, k, s, t) corresponding to The outbound station s and outbound time t can be used as the predicted destination station, and the outbound time t can be used as the predicted arrival time.

In some embodiments, after the predicted arrival time is obtained, the arrival reminder may also be performed according to the arrival time. As a way, it may be based on the predicted arrival time to trigger the mobile terminal to output prompt information, which is used to remind the user that the user has arrived or is about to arrive. Among them, the prompt information may include the name and estimated time of the station that has arrived or is about to arrive. It can be output by voice or pop-up in the form of a window, which is not limited here.

Step S260: Obtain a second probability corresponding to the current station entry according to the historical ride record, where the second probability is the probability of entering the current station entry under different historical entry times.

Since the actual current station is not the same as the predicted target historical station, it can be considered that the current station time is not suitable for participating in the prediction of the target station based on habits. Therefore, the destination station can be predicted directly based on the historical ride records and the current station entry, ensuring the accuracy of the prediction result. Optionally, the current incoming station is different from the target historical incoming station, which can be expressed by a mathematical formula: InStation! = TEStation.

In some embodiments, the second probability corresponding to the current stop can be obtained according to the historical ride record, so as to subsequently predict the destination stop based on the second probability. Among them, the second probability is the probability of entering the current station at different historical station time. As a way, you can obtain the statistics of the number of historical ride records entering the current station at different historical pit stops, and calculate the ratio of the number of each record to the total number of historical records. The second probability of entering the current station at the historical station time.

Step S270: Obtain the historical entry time corresponding to the second probability with the largest value as the predicted target entry time.

After obtaining the second probability of entering the current station under different historical pitting times, the historical pitting time corresponding to the second probability with the largest value can be used as the target pitting time that can participate in the prediction of the destination station. It is understandable that the probability of entering the current station at a certain pit-stop time is the greatest, which indicates that the user is likely to habitually enter the current station at that pit-stop time. Therefore, you can set the highest value The historical pit-stop time corresponding to the second probability is used as the target pit-stop time predicted by the participating destination site.

In some embodiments, when dividing all historical inbound times into time periods, it may also be to obtain the second probability of entering the current station in different inbound time periods, and then obtain the second probability corresponding to the largest value. The pit-stop time period is used as the predicted target pit-stop time period. As a method, you can obtain the statistics of the number of historical rides entering the current station during different time periods of inbound stations, and calculate the ratio of the number of each record to the total number of historical records. The second probability of entering the current station during the pit stop time period.

In some embodiments, the target entry time with the highest probability can also be directly obtained through the following calculation formula:

Among them, c is substituted into the current station, CEStation is the predicted target station time, p(c,k) is used to calculate the probability of each time period when the station is c, k=1, 2, 3... , KI. That is, p (incoming stop = c and the incoming time class is k).

It is used to obtain the inbound time category k corresponding to the maximum probability p(c,k).

Step S280: Obtain a third probability corresponding to the current inbound station according to the historical ride record, where the third probability is the number of times from the current inbound station to each historical outbound station at the target inbound time Probability.

Step S290: Obtain the historical outbound site corresponding to the third probability with the largest value as the predicted destination site.

After obtaining the most probable target pit stop time based on the current approach station, the destination station can be predicted based on historical ride records, target pit station time, and current approach station. Specifically, you can first obtain the third probability corresponding to the current inbound station according to the historical ride records. The third probability is the probability of going from the current inbound station to each historical outbound station at the target inbound time, and then obtain the highest numerical value. The historical outbound site corresponding to the third probability is used as the predicted destination site. Among them, to obtain the third probability corresponding to the current entry site, and to obtain the historical exit site corresponding to the third probability with the largest value, please refer to the aforementioned acquisition of the first probability corresponding to the current entry site and the historical exit site corresponding to the first probability with the largest value. The relevant content of the site will not be repeated here.

In some embodiments, the current arrival time can also be predicted. Specifically, when the third probability of going from the current inbound station to each historical outbound station under the predicted target inbound time is obtained, the historical outbound time corresponding to the third probability with the largest value can be used as the predicted Arrival time.

At present, it is also possible to substitute the inbound time period class corresponding to the target inbound time and the current inbound station into c and k in formula (4), so that the outbound corresponding to the maximum probability p(c, k, s, t) can be obtained. Station s and outbound time t, the outbound site can be used as the predicted destination site, and the outbound time t can be used as the predicted arrival time. In some embodiments, after the predicted arrival time is obtained, the mobile terminal can also be triggered to output prompt information according to the arrival time to remind the user that the station has arrived or is about to arrive.

In some embodiments, if there is a prediction result with the same maximum probability, it may indicate that the user's riding habits have changed, and the destination station can be temporarily output as empty. After a certain period is met and the database is more complete, the prediction can be continued. The method for predicting the destination station provided by the embodiment of this application obtains the current station information, which can include the current station time and the current station, and then obtains the target that matches the current station time according to the historical ride records Historical entry site, to predict the destination site based on historical ride records, current entry time, and current entry site when the current entry site is the same as the target historical entry site; and when the current entry site is different from the target historical entry site , According to the historical ride records and the current station, the pit time can be predicted to obtain the predicted target pit time, and then the destination station can be predicted based on the historical ride records, the predicted target pit time and the current station. In this way, taking into account the influence of the current pit stop time on the prediction result, if the forecast of the entry station based on the current pit stop time is accurate, the forecast can be based on the current pit stop time and the current entry station, and based on the current pit stop time In the case that the predicted entry site is not very accurate, it can also be directly predicted based on the current entry site. Improve the accuracy of the destination site prediction, making the predicted destination more in line with the actual needs of users.

Please refer to FIG. 6, which shows a schematic flowchart of a method for predicting a destination site according to another embodiment of the present application. The following will elaborate on the process shown in FIG. 6, and the prediction method of the destination site shown may specifically include the following steps:

Step S310: Obtain a first destination station predicted based on a first historical ride record, the first historical ride record being a historical ride record dating back to the first date from the current date.

Step S320: Obtain a second destination station predicted based on a second historical ride record, the second historical ride record being a historical ride record retroactive from the current date to a second date, and the second date is earlier than the first A date.

In some embodiments, the destination site can be predicted separately through learning of recent ride records and long-term ride records, so as to improve the accuracy of prediction. Specifically, it is possible to obtain the first destination station predicted based on the first historical ride record, which is the historical ride record dating back to the first date from the current date, and obtain the prediction based on the second historical ride record The second destination site, the second historical ride record is the historical ride record dating back to the second date from the current date. The second date is earlier than the first date, that is, the first historical ride record can be understood as a recent ride record, and the second historical ride record can be understood as a long-term ride record.

It is understandable that the first destination site predicted based on the first historical ride record and the second destination site predicted based on the second historical ride record can be predicted based on the aforementioned prediction method, and will not be repeated here.

It should be noted that the specific retrospective duration of the first historical ride record and the second historical ride record is not limited in this application, and can be set reasonably according to specific scenarios. For example, the retrospective duration of the first historical ride record may be within 48 hours, and the retrospective duration of the second historical ride record may be within half a year or within one year.

Step S330: Determine a predicted destination site according to the first destination site and the second destination site.

After obtaining the first destination site and the second destination site predicted by the recent ride records and long-term ride records, site fusion processing can be performed to obtain the final destination station prediction results.

In some embodiments, different weights may be assigned to the short-term habit prediction results and the long-term habit prediction results, so as to determine the final predicted destination site according to the weights. Specifically, referring to FIG. 7, step S330 may include:

Step S331: Obtain the first weight corresponding to the first destination site and the second weight corresponding to the second destination site, respectively.

Step S332: Obtain a first product of the predicted probability corresponding to the first destination site and the first weight.

Step S333: Obtain a second product of the predicted probability corresponding to the second destination site and the second weight.

Step S334: From the first product and the second product, obtain the destination site corresponding to the product with the largest value as the predicted destination site.

In some embodiments, after the first destination site and the second destination site are acquired, the corresponding first weight and second weight can be acquired respectively. Then the first product of the predicted probability corresponding to the first destination site and the first weight can be calculated, and the second product of the predicted probability corresponding to the second destination site and the second weight can be calculated to obtain from the first product and the second product The destination site corresponding to the product with the largest value is used as the predicted destination site. Among them, the predicted probability corresponding to the first destination site and the predicted probability corresponding to the second destination site can be the maximum probability corresponding to the destination site predicted by the foregoing prediction method, namely

In some embodiments, in order to make the prediction result more consistent with the current user's travel habits, the weight value corresponding to the recent habit prediction result may be higher than the weight value corresponding to the long-term habit prediction result, that is, the first weight is greater than the second weight. Of course, the specific weight setting is not limited here, and it can be set reasonably according to the specific forecast demand.

The method for predicting the destination station provided in the embodiment of the application uses the aforementioned target station prediction method to predict the first destination station based on long-term historical ride records, and uses the aforementioned target station prediction method to predict the first destination station based on recent historical ride records. Second destination site, and then determine the final predicted destination site based on the first destination site and the second destination site. In this way, when predicting the destination site, both the long-term riding habits of the user and the recent riding habits of the user are taken into consideration. When the recent riding habits change, the inaccurate prediction caused by the single learning long-term riding habits can be avoided. , Improve the accuracy of the destination site prediction, making the predicted destination more suitable for the actual needs of users.

In some embodiments, after the predicted destination site is obtained, the arrival reminder function can also be implemented. Specifically, please refer to FIG. 8, which shows a schematic flowchart of a method for predicting a destination site according to another embodiment of the present application. The following will elaborate on the process shown in FIG. 8. The prediction method of the destination site shown may specifically include the following steps:

Step S410: Determine whether the current site is a neighboring site of the destination site. If yes, step S420 can be executed; if not, step S430 can be executed.

Step S420: Trigger the mobile terminal to output prompt information, which is used to remind the user that the station is about to arrive.

Step S430: The step of triggering the mobile terminal to output prompt information is not executed, and the prompt information is used to remind the user that the station is about to arrive.

In the embodiment of the present application, the current site can be located in real time, so as to trigger the arrival reminder function when the current site reaches the neighboring site of the destination site. Specifically, when the current site is acquired, it can be determined whether the current site is a neighboring site of the destination site. Such as the site before the destination site. When the current site is an adjacent site of the destination site, the mobile terminal can be triggered to output prompt information, which is used to remind the user that the station is about to arrive. When the current site is not an adjacent site of the destination site, the mobile terminal may not be triggered to output prompt information. As shown in Figure 9, Figure 9 shows a schematic diagram of an application scenario where a user enters a station from A and predicts a destination site E. The mobile phone perceives A->B->C->D and reminds the user to arrive when it is approaching E. Among them, the current site can be located in multiple ways such as GPS, WiFi, Bluetooth, base station, and IMU, which is not limited here.

In some embodiments, the current site can be located based on the base station information. Specifically, referring to FIG. 10, before step S410, the method for predicting the target site of the present application may further include:

Step S400: Obtain the information of the base station to which the mobile terminal is currently connected.

Generally, mobile terminals such as mobile phones of users are plugged into mobile phone cards, because the mobile phone is connected to the base station of China Mobile, China Unicom, or telecom operator, so that the mobile phone can make calls and surf the Internet. Therefore, in some embodiments, information of the base station to which the mobile terminal is currently connected can be obtained. The base station information includes the unique Cell ID of the base station, the operator, and other information, so as to query the database based on the base station information.

Step S401: Acquire a target site corresponding to the base station information according to the corresponding relationship between the base station and the site stored in the base station database.

For example, in a subway scene, as shown in FIG. 11, FIG. 11 shows a schematic diagram of base station coverage. It can be seen that the coverage of the base station is limited. The coverage range of the base station is usually about 2 kilometers, and generally no more than 5 kilometers. Taking into account the poor signal quality in subway stations, major operators will deploy more base stations in the subway to ensure the quality of calls and Internet access on the subway. Therefore, at different subway stations, mobile phones will be registered to unreachable base stations.

In some embodiments, a base station and site mapping table can be established and stored in the database to obtain the base station database. Among them, the content of the database table can be obtained by manually going offline to the actual site and scanning the surrounding base station information with a mobile phone. Therefore, the target site corresponding to the information of the base station currently connected to the mobile terminal can be obtained according to the corresponding relationship between the base station and the site stored in the base station database.

For example, the established mapping table between the base station and the subway station may be as shown in Table 2. According to the information of the base station to which the mobile terminal is currently connected, query the database to obtain the subway stations covered by the base station. So you can locate the user's current site.

Table 2

Base station information	Subway station information
mcc-460-mnc-00-ci-243174306-pci-303-tac-6227	Shanghai Line 7-Longhua Middle Road
mcc-460-mnc-01-ci-11114774-pci-501-tac-6263	Shanghai Line-12-Longcao Road
mcc-460-mnc-01-ci-11115275-pci-466-tac-6263	Shanghai Line-12-Longhua Middle Road

In some embodiments, data collection may be incomplete, or the operator may add a new base station during subsequent maintenance, resulting in the base station database not being queried for corresponding base station information, and the target site may not be obtained. At this time, other positioning methods can be used to locate the current site.

Step S402: Determine the current site according to the target site.

In some embodiments, when the target site corresponding to the base station information is obtained, and the corresponding target site is one, the target site may be used as the current site. It can indicate that the user is about to arrive at the site or is currently at the site at this time.

In some embodiments, when the coverage of the base station is relatively wide, the information of the base station may cover 2 or more sites. In this case, it is determined which site the user is currently at. Therefore, the previous site needs to be used to assist in the determination. Specifically, when there are multiple target sites, step S402 may include:

Step S4021: Obtain neighboring sites of the previous site.

Step S4022: Obtain multiple identical sites among the target sites and the neighboring sites as the current site.

In some embodiments, when the current site needs to be located, the previously determined site can be acquired. In this way, the neighboring site of the previous station can be obtained according to the official route information, and then multiple target sites covered by the base station and the same site in the neighboring site can be obtained as the current site. Exemplarily, please refer to FIG. 12, which shows a schematic diagram of an overall flow of base station positioning. It should be noted that at the initial moment of the ride, the determined stop is the pit stop.

For example, at the ABCDE subway station, if the current station is determined to be B (also known as the previous station), since the subway is always moving forward, the current station needs to be determined in real time, and the adjacent stations of B are set (A, C), if The currently connected base station covers 2 subway stations (C, D), the intersection of the two sets is C, and C is the current station after the subway travels (that is, the next station after B).

As a specific embodiment, in the subway scene, please refer to FIG. 13, which shows a schematic block diagram of the overall flow of prediction of a destination site. in:

Inbound recognition module 101: the user enters the subway station to recognize, the methods include mobile phone swiping recognition, mobile phone NFC recognition, inbound location GPS recognition, subway driving recognition, this module is embedded in the android system. When a user uses a mobile phone APP to swipe the code to enter a subway station, for example, the Shanghai subway is a metropolitan city, Shenzhen has a WeChat applet Shenzhen subway, and Beijing uses Yitong. After swiping the code to enter the station, the successful entry and the entry site will pop up on the phone. The inbound and outbound recognition module is a part of the customized system, which can identify changes on the mobile phone screen, inbound messages and inbound sites.

The subway riding habit database module 102: a database of the user's subway riding habit. The database is a data collection of the subway riding information of the user for six months or one year.

Module 103 of subway ride habit within 48 hours: A 48-hour database of subway ride habits of users, which is a data collection of users’ recent subway rides.

The destination site prediction module 104: predicts the user's destination site under the condition of entering the site. For the specific prediction method, refer to the content of the foregoing embodiment.

Target site fusion result module 105: The user's long-term subway ride habit predicts the target site and the 48-hour subway ride habit predicts the target site to merge to obtain the final target station prediction result.

Metro station recognition module 106: an algorithm for station recognition in subway mode, which recognizes the information of the current subway station, such as line, name, etc.

Destination station reminding module 107: Based on the identification of subway mode stations, if the current station and the destination station are adjacent stations, the mobile phone sends notifications and vibrations to remind the user that they are about to arrive at the station and prepare to get off the train.

The method for predicting the destination site provided by the embodiment of the application, after obtaining the predicted destination site through the aforementioned method for predicting the destination site, the current site can be located in real time, and the current site can be triggered when the current site is the neighboring site of the destination site The mobile terminal outputs prompt information, which is used to remind the user that the station is about to arrive. This eliminates the need for user participation, and provides travel users with humanized services such as station arrival reminders and subway operation status, which solves the pain points of users when traveling and prevents users from passing the station. And there is no need for additional costs, no additional deployment equipment, no dependence on the outside, and feasibility and leadership.

Please refer to FIG. 14, which shows a structural block diagram of a destination site prediction device 700 provided by an embodiment of the present application. The destination site prediction device 700 includes: a current inbound acquisition module 710, a target inbound acquisition module 720, The first prediction module 730 and the second prediction module 740. Wherein, the current inbound acquisition module 710 is used to acquire current inbound information, and the current inbound information includes the current inbound time and the current inbound station; the target inbound acquisition module 720 is used to obtain information related to the current inbound station based on historical ride records. The target historical entry point with the pit stop time matching; the first prediction module 730 is used for when the current entry point is the same as the target historical entry point, according to the historical ride records, the current pit stop time and the current The inbound station predicts the destination station; the second prediction module 740 is used to predict the destination station based on the historical ride records and the current inbound station when the current inbound station is different from the target historical inbound station.

In some embodiments, the first prediction module 730 may include: a first probability acquisition unit and a first probability prediction unit. Wherein, the first probability obtaining unit is configured to obtain a first probability corresponding to the current arrival station according to the historical ride record, and the first probability is that the current arrival station is to go from the current station at the current arrival time. The probability of each historical exit site; the first probability prediction unit is used to obtain the historical exit site corresponding to the first probability with the largest value, as a predicted destination site.

In some embodiments, the first probability obtaining unit may be specifically configured to: obtain a historical inbound time matching the current inbound time according to the historical ride record; obtain the historical inbound time and different historical outbound times. At station time, the transition probability from the current inbound station to each historical outbound station is taken as the first probability corresponding to the current inbound station.

In some embodiments, the prediction device 700 of the destination site may further include: an arrival time prediction module and a prompt triggering module. Wherein, the arrival time prediction module is used to obtain the historical outbound time corresponding to the first probability with the largest value as the predicted arrival time; the prompt trigger module is used to trigger the mobile terminal to output prompt information according to the arrival time , The prompt information is used to remind the user that he has arrived at the station or is about to arrive at the station.

In some embodiments, the second prediction module 740 may include: a second probability acquisition unit, a station time prediction unit, a third probability acquisition unit, and a third probability prediction unit. Wherein, the second probability obtaining unit is configured to obtain a second probability corresponding to the current station entry according to the historical ride record, and the second probability is the probability of entering the current station station at different historical pit stops. The pit stop time prediction unit is used to obtain the historical pit stop time corresponding to the second probability with the largest value as the predicted target pit stop time; the third probability acquisition unit is used to obtain the current pit stop time according to the historical ride record The third probability corresponding to the entry site, the third probability is the probability of going from the current entry site to each historical exit site under the target entry time; the third probability prediction unit is used to obtain the highest numerical value The historical outbound site corresponding to the third probability is used as the predicted destination site.

In some embodiments, the device 700 for predicting a destination site may further include: a first site acquisition module, a second site acquisition module, and a destination site prediction module. Wherein, the first site acquisition module is used to acquire the first destination site predicted based on the first historical ride record, the first historical ride record being the historical ride record dating back to the first date from the current date; the second site acquires The module is used to obtain a second destination station predicted based on a second historical ride record, the second historical ride record being a historical ride record retroactive from the current date to a second date, the second date being earlier than the first A date; the destination site prediction module is used to determine the predicted destination site based on the first destination site and the second destination site.

In some embodiments, the aforementioned destination site prediction module may be specifically configured to: obtain the first weight corresponding to the first destination site and the second weight corresponding to the second destination site respectively; and obtain the corresponding first destination site. The first product of the predicted probability of and the first weight; obtain the second product of the predicted probability of the second destination site and the second weight; from the first product and the second product, Obtain the destination site corresponding to the product with the largest value as the predicted destination site.

In some embodiments, the device 700 for predicting the destination site may further include: a base station information acquisition module, a target site acquisition module, and a current site positioning module. Among them, the base station information acquisition module is used to acquire the base station information that the mobile terminal is currently connected to; the target site acquisition module is used to acquire the target site corresponding to the base station information according to the corresponding relationship between the base station and the site stored in the base station database; the current site location The module is used to determine the current site according to the target site.

In some embodiments, the above-mentioned current station positioning module may be specifically used to: obtain a neighboring station of the previous station; and obtain the same station among multiple target stations and the neighboring stations as the current station.

Those skilled in the art can clearly understand that, for the convenience and conciseness of description, the specific working process of the device and module described above can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

In the several embodiments provided in this application, the coupling between the modules may be electrical, mechanical or other forms of coupling.

In addition, the functional modules in the various embodiments of the present application may be integrated into one processing module, or each module may exist alone physically, or two or more modules may be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or software functional modules.

In summary, the prediction device of the destination site provided in the embodiment of the present application is used to implement the prediction method of the corresponding destination site in the foregoing method embodiment, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

Please refer to FIG. 15, which shows a structural block diagram of an electronic device provided by an embodiment of the present application. The electronic device 100 may be a mobile terminal capable of running application programs, such as a notebook computer, a smart phone, a smart watch, or a smart glasses. The electronic device 100 in this application may include one or more of the following components: a processor 110, a memory 120, and one or more application programs, where one or more application programs may be stored in the memory 120 and configured to be configured by One or more processors 110 execute, and one or more application programs are configured to execute the methods described in the foregoing method embodiments.

The processor 110 may include one or more processing cores. The processor 110 uses various interfaces and lines to connect various parts of the entire electronic device 100, and executes by running or executing instructions, programs, code sets, or instruction sets stored in the memory 120, and calling data stored in the memory 120. Various functions and processing data of the electronic device 100. Optionally, the processor 110 may adopt at least one of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). A kind of hardware form to realize. The processor 110 may integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a predictor (Graphics Processing Unit, GPU) of a destination site, and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is used for rendering and drawing of display content; the modem is used for processing wireless communication. It can be understood that the above-mentioned modem may not be integrated into the processor 110, but may be implemented by a communication chip alone.

The memory 120 may include random access memory (RAM) or read-only memory (Read-Only Memory). The memory 120 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 120 may include a program storage area and a data storage area, where the program storage area may store instructions for implementing the operating system and instructions for implementing at least one function (such as touch function, sound playback function, image playback function, etc.) , Instructions used to implement the following various method embodiments, etc. The storage data area can also store data (such as phone book, audio and video data, chat record data) created by the electronic device 100 during use.

It can be understood that the structure shown in FIG. 15 is only an example, and the electronic device 100 may also include more or fewer components than those shown in FIG. 15, or have a completely different configuration from that shown in FIG. 15. The embodiments of the present application have no limitation on this.

Please refer to FIG. 16, which shows a structural block diagram of a computer-readable storage medium provided by an embodiment of the present application. The computer-readable medium 800 stores program code, and the program code can be invoked by a processor to execute the method described in the foregoing method embodiment.

The computer-readable storage medium 800 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory), EPROM, hard disk, or ROM. Optionally, the computer-readable storage medium 800 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 800 has storage space for the program code 810 for executing any method steps in the above-mentioned methods. These program codes can be read from or written into one or more computer program products. The program code 810 may be compressed in a suitable form, for example.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the application, not to limit them; although the application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: The technical solutions recorded in the foregoing embodiments are modified, or some of the technical features thereof are equivalently replaced; these modifications or replacements do not drive the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims (20)

1. A method for predicting a destination site, characterized in that the method includes:

   Acquire current inbound information, where the current inbound information includes the current inbound time and the current inbound station;

   According to the historical ride records, obtain the target historical entry site matching the current entry time;

   When the current inbound station is the same as the target historical inbound station, predict the destination station based on the historical ride records, the current inbound time, and the current inbound station;

   When the inbound station is different from the target historical inbound station, the destination station is predicted based on the historical ride records and the current inbound station.
2. The method according to claim 1, wherein the predicting a destination station based on the historical ride record, the current arrival time, and the current arrival station comprises:

   Obtaining, according to the historical ride record, a first probability corresponding to the current inbound station, where the first probability is the probability of going from the current inbound station to each historical outbound station at the current inbound time;

   Obtain the historical outbound site corresponding to the first probability with the largest value as the predicted destination site.
3. The method according to claim 2, wherein the obtaining the first probability corresponding to the current stoppage according to the historical ride record comprises:

   According to the historical ride records, obtain a historical pit stop time that matches the current pit stop time;

   Obtain the transition probability from the current inbound station to each historical outbound station under the historical inbound time and different historical outbound times as the first probability corresponding to the current inbound station.
4. The method according to claim 2 or 3, wherein the method further comprises:

   Acquiring the historical outbound time corresponding to the first probability with the largest value as the predicted arrival time;

   According to the arrival time, the mobile terminal is triggered to output prompt information, and the prompt information is used to remind the user that the station has arrived or is about to arrive.
5. The method according to any one of claims 1 to 4, wherein the predicting a destination station based on the historical ride record and the current arrival station comprises:

   According to the historical ride records, obtain a second probability corresponding to the current stop, where the second probability is the probability of entering the current stop at a different historical stop time;

   Acquiring the historical inbound time corresponding to the second probability with the largest value as the predicted target inbound time;

   According to the historical ride records, obtain a third probability corresponding to the current inbound station, where the third probability is the probability of going from the current inbound station to each historical outbound station at the target inbound time;

   Obtain the historical outbound site corresponding to the third probability with the largest value as the predicted destination site.
6. The method according to any one of claims 1-5, wherein the method further comprises:

   Acquiring a first destination station predicted based on a first historical ride record, the first historical ride record being a historical ride record dating back to the first date from the current date;

   Acquiring a second destination site predicted based on a second historical ride record, the second historical ride record being a historical ride record retroactive from the current date to a second date, the second date being earlier than the first date;

   Determine the predicted destination site according to the first destination site and the second destination site.
7. The method according to claim 6, wherein the determining the predicted destination site according to the first destination site and the second destination site comprises:

   Acquiring the first weight corresponding to the first destination site and the second weight corresponding to the second destination site respectively;

   Acquiring a first product of the predicted probability corresponding to the first destination site and the first weight;

   Acquiring a second product of the predicted probability corresponding to the second destination site and the second weight;

   From the first product and the second product, obtain the destination site corresponding to the product with the largest value as the predicted destination site.
8. 8. The method of claim 7, wherein the first weight is greater than the second weight.
9. The method according to any one of claims 1-8, wherein the acquiring, based on historical ride records, a target historical entry point matching the current entry time comprises:

   According to the historical ride records, obtain the inbound time period that matches the current inbound time, where the inbound time period is obtained by dividing all historical inbound times in the historical ride records into sections;

   According to the historical ride records, obtain the entry probability of entering different historical entry stations during the said entry time period;

   Obtain the historical entry site corresponding to the entry probability with the largest value as the target historical entry site.
10. The method according to claim 9, characterized in that the obtaining the pit stop time period matching the current pit stop time according to historical ride records comprises:

    Obtain all historical pit stops according to historical ride records;

    Perform a cluster analysis on all the historical inbound times to obtain multiple time periods after clustering;

    The time period closest to the current inbound time is acquired from the multiple time periods as the inbound time period that matches the current inbound time.
11. The method according to claim 10, wherein the time period closest to the current inbound time is obtained from the multiple time periods as the inbound time period that matches the current inbound time ,include:

    Acquiring the prior probability of each time period in the multiple time periods;

    From the multiple time periods, obtain a target time period in which the prior probability is greater than a specified threshold;

    The time period closest to the current inbound time is acquired from the target time period as the inbound time period that matches the current inbound time.
12. The method according to any one of claims 9-11, wherein the predicting a destination station based on the historical ride record, the current arrival time, and the current arrival station comprises:

    According to the historical ride records, obtain the transition probability from the current inbound station to each historical outbound station in the inbound time period and different outbound time periods as the first probability corresponding to the current inbound station , Wherein, the outbound time period is a time period obtained by dividing all historical outbound times in the historical ride records into sections;

    Obtain the historical outbound site corresponding to the first probability with the largest value as the predicted destination site.
13. The method according to claim 12, wherein the method further comprises:

    Obtain all historical departure times according to historical ride records;

    Perform cluster analysis on all the historical outbound times to obtain multiple outbound time periods after clustering.
14. The method according to any one of claims 1-13, wherein the method further comprises:

    When the current site is an adjacent site of the destination site, the mobile terminal is triggered to output prompt information, and the prompt information is used to remind the user that the station is about to arrive.
15. The method according to claim 14, characterized in that, before triggering the mobile terminal to output prompt information when the current site is a neighboring site of the destination site, the method further comprises:

    Acquiring information about the base station to which the mobile terminal is currently connected;

    Acquiring the target site corresponding to the base station information according to the corresponding relationship between the base station and the site stored in the base station database;

    According to the target site, the current site is determined.
16. The method according to claim 15, wherein there are multiple target sites, and the determining the current site according to the target site comprises:

    Get the neighboring site of the previous site;

    Acquire multiple identical sites among the target site and the neighboring sites as the current site.
17. The method according to any one of claims 1-16, wherein the historical ride record includes historical inbound information and historical outbound information, wherein the historical inbound information includes historical inbound station and historical inbound station Time, the historical outbound information includes historical outbound site and historical outbound time.
18. A prediction device for a destination site, characterized in that the device includes:

    The current inbound acquisition module is used to obtain current inbound information, and the current inbound information includes the current inbound time and the current inbound station;

    The target inbound acquisition module is used to obtain the target historical inbound stop matching the current inbound time according to the historical ride records;

    The first prediction module is configured to predict the destination station based on the historical ride record, the current pit stop time, and the current pit stop when the current inbound station is the same as the target historical inbound station;

    The second prediction module is used to predict the destination station based on the historical ride records and the current arrival station when the current inbound station is different from the target historical inbound station.
19. An electronic device, characterized in that it comprises:

    One or more processors;

    Memory

    One or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute such as The method of any one of claims 1-17.
20. A computer-readable storage medium, wherein the computer-readable storage medium stores program code, and the program code can be called by a processor to execute the method according to any one of claims 1-17 .

Images