A travel mode recommendation method, related method and device, and system

Abstract

The application discloses a travel mode recommendation method, related method, device and system. The travel mode recommendation method comprises the following steps: obtaining route planning data of a travel object in response to a route planning request, wherein the route planning data at least comprises a starting point of the object, a terminal point of the object, a travel time of the object and a distance between the starting point and the terminal point of the object; obtaining historical route planning data of the travel object, wherein the historical route planning data at least comprises a historical travel mode, a historical distance between the starting point and the terminal point and a historical travel time; generating feature data based on the historical route planning data and the route planning data; and inputting the feature data into a pre-trained travel mode prediction model to predict a travel mode recommended to the travel object. The method improves the accuracy of travel mode prediction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a travel mode recommendation method, related method, device and system. BACKGROUND

[0002] Navigation service is a service frequently used by travel objects traveling with the help of travel applications (such as map navigation applications or online car-hailing applications, etc.). Navigation service is roughly divided into three stages:

[0003] 1) Route request stage: the travel object initiates a route planning request through the travel application;

[0004] 2) Route planning stage: the navigation service plans a route according to the received route planning request, and returns a plurality of recommended planning routes to the travel application for the travel object to select;

[0005] 3) Navigation guidance stage: the travel object selects one from the recommended planning routes through the travel application and starts the navigation guidance service.

[0006] The present application is found by the inventor during the research on the above-mentioned technology that if the navigation service can accurately predict the travel mode (driving, taking the bus, riding, walking, etc.) required by the travel object before route planning, and plan and recommend a route based on the travel mode required by the travel object, on the one hand, it can bring good user experience to the travel object, and on the other hand, the travel mode recommendation is accurate, which can significantly reduce the consumption of the computing resources of the server, because if the travel mode prediction is inaccurate, the travel object needs to switch the travel mode, and the navigation server also needs to plan a route again according to the travel mode after the switching, for example, the predicted travel mode is driving, the navigation service recommends a driving planning route to the user, and the actual travel mode required by the travel object is taking the bus and subway, so the navigation service needs to plan a bus and subway route for the travel object, and in this case, the computing resources consumed by the navigation server for driving route planning are wasted. Therefore, how to improve the accuracy of travel mode prediction is a technical problem that needs to be continuously solved and optimized in the field. SUMMARY

[0007] In view of the above problems, the present application is proposed to provide a travel mode recommendation method, related method, device and system which can overcome the above problems or at least partially solve the above problems.

[0008] In a first aspect, the embodiments of the present application provide a travel mode recommendation method, comprising:

[0009] In response to the route planning request, route planning data of the travel object is obtained, the route planning data at least including: object starting point, object ending point, object travel time and object starting and ending point distance;

[0010] The historical route planning data of the travel object is obtained, the historical route planning data at least including: historical travel mode, historical starting and ending point distance and historical travel time;

[0011] The feature data is generated based on the historical route planning data and the route planning data;

[0012] The feature data is input into a pre-trained travel mode prediction model, and a travel mode recommended to the travel object is predicted.

[0013] In a second aspect, an embodiment of the present application provides a training method of a travel mode prediction model, comprising:

[0014] Route planning data corresponding to a historical route planning request as a sample is obtained as sample route planning data, the sample route planning data at least including: object starting point, object ending point, object travel mode, object travel time and object starting and ending point distance;

[0015] Route planning data corresponding to a historical route planning request with historical travel time earlier than the object travel time in the sample route planning data is obtained as historical route planning data, the historical route planning data at least including: historical travel mode, historical starting and ending point distance and historical travel time;

[0016] The feature data for model training is obtained based on the sample route planning data and the historical route planning data;

[0017] The feature data is taken as input of a travel mode prediction model to be trained, and the model is trained until parameters of the model meet requirements according to the object travel mode in the sample route planning data.

[0018] In a third aspect, an embodiment of the present application provides a travel mode recommendation device, comprising:

[0019] A first information obtaining module is configured to obtain route planning data of a travel object in response to a route planning request, the route planning data at least including: starting point, ending point, current request time and starting and ending point distance;

[0020] A second information obtaining module is configured to obtain historical route planning data of the travel object, the historical route planning data at least including: historical travel mode, historical starting and ending point distance and historical request time;

[0021] The learning feature determination module is configured to obtain feature data for model learning based on the historical route planning data and the route planning data.

[0022] The prediction module is configured to input the feature data into a pre-trained travel mode prediction model to predict a travel mode recommended for a travel subject.

[0023] In a fourth aspect, an embodiment of the present application provides a training device of a travel mode prediction model, comprising:

[0024] The third information acquisition module is configured to acquire route planning data corresponding to a historical route planning request as sample route planning data, the sample route planning data at least including a subject starting point, a subject ending point, a subject travel mode, a subject travel time and a subject starting and ending point distance.

[0025] The fourth information acquisition module is configured to acquire route planning data corresponding to a historical route planning request with a historical travel time earlier than the subject travel time in the sample route planning data as historical route planning data, the historical route planning data at least including a historical travel mode, a historical starting and ending point distance and a historical travel time.

[0026] The training feature determination module is configured to obtain feature data for model training based on the sample route planning data and the historical route planning data.

[0027] The model training module is configured to input the feature data as an input of a travel mode prediction model to be trained, train the model until parameters of the model meet requirements according to the subject travel mode in the sample route planning data.

[0028] In a fifth aspect, an embodiment of the present application provides a travel mode recommendation system, comprising a navigation server and at least one terminal device, wherein:

[0029] The navigation server is provided with the travel mode recommendation device, configured to receive a route planning request sent by the terminal device, and send a determined travel mode to the terminal device.

[0030] The terminal device is configured to send the route planning request to the navigation server, and receive a travel mode returned by the navigation server.

[0031] In a sixth aspect, an embodiment of the present application provides a travel mode recommendation service, the service being configured to execute the travel mode recommendation method.

[0032] In a seventh aspect, an embodiment of the present application provides a model training service, the service being configured to execute the training method of the travel mode prediction model.

[0033] The beneficial effects of the above-mentioned technical solutions provided in the embodiments of this application include at least the following:

[0034] The travel mode recommendation method, related methods, apparatus, and system provided in this application embodiment obtain route planning data and historical route planning data of the traveler. Based on the route planning data and historical route planning data, feature data is obtained. The feature data is used to predict and recommend the travel mode of the traveler through a pre-trained travel mode prediction model. Since the feature data integrates the traveler's historical data (historical route planning data) and real-time data (route planning data in the route planning request), and the travel mode prediction model better predicts the travel mode, the accuracy of travel mode prediction is improved. On the one hand, it reduces the number of times the traveler switches between travel modes during route planning, thereby reducing the complexity of the traveler's route planning operation, saving the traveler's route planning time, improving the user experience of the traveler, and reducing the route calculation cost of the navigation server.

[0035] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0036] The technical solution of this application will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0037] The accompanying drawings are provided to further illustrate the present application and form part of the specification. They are used together with the embodiments of the present application to explain the application and do not constitute a limitation thereof. In the drawings:

[0038] Figure 1 A flowchart illustrating the travel mode recommendation method provided in this application embodiment;

[0039] Figure 2 A flowchart illustrating the training method for the travel pattern prediction model provided in this application embodiment;

[0040] Figure 3a A schematic diagram illustrating the process of obtaining historical real route planning requests in the travel mode recommendation method provided in this application embodiment;

[0041] Figure 3b for Figure 3a A schematic diagram illustrating the process for determining historical route planning requests;

[0042] Figure 4A flowchart of a process of obtaining a historical route planning request as a sample in a training method of a travel mode prediction model provided by an embodiment of the present application is shown.

[0043] Figure 5 A structural diagram of a navigation service system provided by an embodiment of the present application is shown.

[0044] Figures 6a-6c A flowchart of a process of implementing route planning by a map navigation application in a terminal device provided by an embodiment of the present application is shown.

[0045] Figure 7 A structural diagram of a travel mode recommendation device provided by an embodiment of the present application is shown.

[0046] Figure 8 A structural diagram of a training device of a travel mode prediction model provided by an embodiment of the present application is shown.

[0047] Figure 9 A structural diagram of a travel mode recommendation system provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0048] Exemplary embodiments of the present application will be described in detail with reference to the drawings. Although exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present application can be more thoroughly understood, and the scope of the present application can be accurately conveyed to those skilled in the art.

[0049] Regarding travel mode recommendation, the existing technology usually recommends the travel mode used by a travel object in the last time, i.e., the travel mode used in the last time closest to the current travel time of the travel object is taken as the current travel mode of the travel object, which is prone to have low accuracy.

[0050] To solve the problems of the existing technology, the present application provides a travel mode recommendation method to improve the accuracy of travel mode prediction, reduce the complexity of route planning operation of a travel object, save the route planning time of the travel object, improve the user experience of the travel object, and reduce the cost of a navigation server.

[0051] The technical solutions provided by the embodiments of the present application will be described below through specific embodiments.

[0052] The specific implementation of the travel mode recommendation method, related methods, devices and systems provided by the embodiments of the present application will be described in detail below.

[0053] Embodiment One

[0054] The embodiment of the application provides a travel mode recommendation method, and the flow thereof is shown in the figure and comprises the following steps: Figure 1

[0055] S101, in response to a route planning request, obtaining route planning data of a travel object, the route planning data at least comprising: object starting point, object ending point, object travel time and object starting and ending point distance;

[0056] The object starting and ending point distance can be the straight-line distance between the object starting point and the object ending point. The object starting point and the object ending point can be the latitude and longitude position of the object starting point and the latitude and longitude position of the object ending point, wherein the latitude and longitude position of the object starting point can be obtained by positioning, and the positioning mode can adopt the mode in the prior art, for example, GPS positioning, base station positioning, WIFI assisted positioning, etc. The latitude and longitude position of the object ending point is generally determined in an electronic map by the selected ending point of the travel object.

[0057] The object travel time can be the travel time selected by the travel object when initiating the route planning request, for example, the travel object initiates the route planning request through a map navigation App at 15:00 in the afternoon, but the travel time selected by the travel object is 15:20, so the travel time is 15:20 at this time; if the travel object does not select the travel time when initiating the route planning request, the time when the travel object initiates the route planning request can be used as the object travel time, for example, the travel object initiates the route planning request through a map navigation App at 15:00 in the afternoon, and the travel object does not select the travel time, so the travel time is 15:00 at this time.

[0058] S102, obtaining historical route planning data of the travel object, the historical route planning data at least comprising: historical travel mode, historical starting and ending point distance and historical travel time;

[0059] The historical route planning data of the travel object can be obtained by the following mode:

[0060] The historical route planning data of the travel object can be obtained by the following mode:

[0061] ​In the historical route planning request of the travel object in a period of time, there may be some route planning requests that cannot represent the real travel demand of the travel object. This is because when performing route planning, the navigation server may recommend an incorrect travel mode, the travel object may misoperate, or the travel object may just want to compare the results of route planning of various travel modes. Therefore, in order to obtain more accurate feature data, the historical route planning data of the travel object obtained in the present scheme can be route planning data corresponding to historical real route planning requests representing the real travel demand of the travel object. The historical real route planning request of the travel object can represent the real travel demand of the user, and therefore, the route planning data corresponding to the historical real route planning request can be used as the historical route planning data of the travel object.

[0062] S103, generating feature data based on the historical route planning data and the route planning data;

[0063] The feature data can be obtained by combining the historical route planning data and the route planning data. The feature data can be a vector, and each information contained in the vector can be represented by one or more parameters in the vector. For example, when the route planning data includes the object starting point, the object ending point, the object travel time, and the object starting and ending point distance, if the object starting point and the object ending point are expressed by the latitude and longitude position, the object starting point is composed of the starting point longitude and the starting point latitude, and the object ending point is composed of the ending point longitude and the ending point latitude, so a 6-dimensional feature can be generated according to the route planning data. Assuming that the historical route planning data includes n historical real route planning requests, where n is a positive integer, and each historical route planning data includes the historical travel mode, the historical starting and ending point distance, and the historical travel time, a 3n-dimensional feature can be generated according to the n historical route planning data. Finally, the feature data obtained by combining the 6-dimensional feature and the 3n-dimensional feature is a vector including a 3n+6-dimensional feature.

[0064] In the embodiments of the present application, in order to make the obtained feature data more convenient for operation and processing in the travel mode prediction model, the parameter of the historical travel time can be replaced by the parameter of the travel time interval between the object travel time in the route planning data and the historical travel time in the historical route planning data, and the step S103 can be implemented by the following steps:

[0065] determining a first travel time interval between the object travel time and the historical travel time;

[0066] generate feature data based on at least the first travel time interval, the historical travel mode, the historical start-end point distance, the object start point, the object end point, the object travel time, and the object start-end point distance.

[0067] Of course, the information of the obtained feature data is richer in the embodiments of the present application. The route planning data can further include other information of the route planning request of the travel object, and the historical route planning data can further include other information of the historical route planning request, for example, point of interest (POI) information, so that the obtained feature data can further include a POI parameter.

[0068] S104, input the feature data into a pre-trained travel mode prediction model to predict a travel mode recommended for the travel object.

[0069] The travel mode prediction model can be trained using a random forest model, a deep neural network model (DNN), a model based on gradient boosting iterative decision trees (GBDT), or other classifier models. The embodiments of the present application do not make specific limitations thereon. Thus, the feature data used for prediction is input into a pre-trained random forest model, a deep neural network model (DNN), a model based on gradient boosting iterative decision trees (GBDT), or other classifier models, to predict a travel mode recommended for the travel object. In the embodiments of the present application, the predicted travel mode of the travel object includes but is not limited to driving, cycling, taking public transportation, and walking modes.

[0070] In one embodiment, the travel mode prediction model described above is trained using route planning data corresponding to historical route planning requests completed by the travel object. The flow thereof is shown in FIG. 2, and specifically includes the following steps: Figure 2

[0071] S201, obtain route planning data corresponding to a historical route planning request as a sample as sample route planning data, the sample route planning data at least including: object start point, object end point, object travel mode, object travel time, and object start-end point distance;

[0072] ​S202, acquire historical route planning data corresponding to historical route planning requests with historical travel times earlier than the object travel times in the sample route planning data as historical route planning data, the historical route planning data at least including: historical travel mode, historical start-end point distance and historical travel time;

[0073] S203, based on the sample route planning data and the historical route planning data, obtain feature data for model training;

[0074] S204, take the feature data as the input of the travel mode prediction model to be trained, train the model until the parameters of the model meet the requirements according to the object travel mode in the sample route planning data.

[0075] In one specific embodiment, when performing the above step S102, historical real route planning requests of travel objects can be pre-screened, and then a preset number of historical real route planning requests are selected from the pre-screened historical real route planning requests of the travel objects in the order from early to late to the object travel time, and further the historical route planning data corresponding to the preset number of historical real route planning requests are acquired as the historical route planning data of the travel objects.

[0076] The process of pre-screening the historical real route planning requests of the travel objects is shown in Figure 3a , and specifically includes the following steps:

[0077] S301: sort the historical route planning requests of the travel objects in the order from early to late to the historical travel time;

[0078] S302: determine the historical real route planning request of the travel object based on the historical travel time, historical travel mode, historical end point and historical start-end point distance in the historical route planning data corresponding to the adjacent two historical route planning requests.

[0079] In one specific embodiment, referring to Figure 3b , the determination of the historical real route planning request of the travel object based on the historical travel time, historical travel mode, historical end point and historical start-end point distance in the historical route planning data corresponding to the adjacent two historical route planning requests in the above step S302 can specifically include the following steps:

[0080] S3021: based on the historical travel time in the historical route planning data corresponding to the adjacent two historical route planning requests, acquire the second travel time interval of the adjacent two historical route planning requests;

[0081] S3022: Obtain whether the two adjacent historical route planning requests are same-trip planning requests based on the historical end point, the historical start-end point distance, and the historical travel mode in the historical route planning data corresponding to the two adjacent historical route planning requests; if yes, perform step S3023; if no, perform step S3024.

[0082] S3023: When the second travel time interval of the two adjacent historical route planning requests is greater than a preset time threshold and the two adjacent historical route planning requests belong to same-trip planning requests, record the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests as the historical real route planning request of the travel object.

[0083] S3024: Obtain the request completion speed of the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests; determine whether the request completion speed is less than a preset speed threshold, if yes, perform step S3025, if no, perform step S3026, the preset speed threshold is related to the historical travel mode in the historical route planning data corresponding to the historical route planning request with earlier historical travel time.

[0084] S3025: When the second travel time interval of the two adjacent historical route planning requests is greater than a preset time threshold, the two adjacent historical route planning requests do not belong to same-trip planning requests, and the request completion speed is less than the preset speed threshold, record the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests as the historical real route planning request of the travel object.

[0085] S3026: Record the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests as the non-historical real route planning request of the travel object.

[0086] With the historical route planning request with later time and another historical route planning request arranged after it in the two adjacent historical route planning requests to form new adjacent historical route planning requests, repeat the foregoing steps S3021-S3026 until it is determined whether the historical route planning request with earlier historical travel time in the last group of two adjacent historical route planning requests is the historical real route planning request of the travel object.

[0087] In one specific embodiment, in step S3022, based on the historical end point, the historical start-end point distance, and the historical travel mode in the historical route planning data corresponding to the two adjacent historical route planning requests, the specific implementation steps of obtaining whether the two adjacent historical route planning requests are same-trip planning requests can be:

[0088] determining whether the historical end point and the historical travel mode in the historical route planning data corresponding to the two adjacent historical route planning requests are the same, and determining whether the historical start-end point distance in the historical route planning data corresponding to the historical route planning request with the later historical travel time is less than the historical start-end point distance in the historical route planning data corresponding to the historical route planning request with the earlier historical travel time;

[0089] If yes, it is determined that the historical route planning request with the later historical travel time in the two adjacent historical route planning requests is the same-trip-in-transit planning request of the historical route planning request with the earlier historical travel time; if no, it is determined that the historical route planning request with the later historical travel time in the two adjacent historical route planning requests is not the same-trip-in-transit planning request of the historical route planning request with the earlier historical travel time.

[0090] In one specific embodiment, in step S3024, the request completion speed of the historical route planning request with the earlier historical travel time in the two adjacent historical route planning requests is obtained, which can be obtained through the following steps:

[0091] The request completion speed of the historical route planning request with the earlier historical travel time is determined based on the historical start-end point distance in the historical route planning data corresponding to the historical route planning request with the earlier historical travel time and the second travel time interval.

[0092] In one specific embodiment, after the historical route planning requests of one travel object are sorted in the order from early to late historical travel time in a selected historical time period, the number of all historical route planning requests of the travel object can be obtained. It is assumed that the historical route planning requests of the travel object are sorted in the order from early to late historical travel time, and 6 historical route planning requests 1-6 are obtained.

[0093] In determining the historical real route planning request of the travel object in the 6 historical route planning requests, first, historical route planning requests 1 and 2 are grouped into a set of adjacent historical route planning requests, and it is determined whether historical route planning request 1 is the historical real route planning request of the travel object through steps S3021-S3026, if yes, the label of historical route planning request 1 is marked as 1, if not, the label of historical route planning request 1 is marked as 0;

[0094] Then, the historical route planning request 2 and 3 are grouped into a set of adjacent historical route planning requests, and the historical route planning request 2 is determined to be a historical real route planning request of the trip object in the same way as that of determining whether the historical route planning request 1 is a historical real route planning request of the trip object, if yes, the label of the historical route planning request 2 is marked as 1, if not, the label of the historical route planning request 2 is marked as 0;

[0095] For the historical route planning requests 3-5, the historical route planning requests 3-5 are sequentially determined to be historical real route planning requests of the trip object in the same way as that of determining whether the historical route planning requests 1 and 2 are historical real route planning requests of the trip object, if yes, the label is marked as 1, if not, the label is marked as 0;

[0096] Finally, after the historical route planning requests 1-5 are all determined to be historical real route planning requests of the trip object, the historical route planning requests 1-5 correspond to a label 0 or 1 respectively, and only the historical route planning requests corresponding to the label 1 in the historical route planning requests 1-5 are screened out, the historical real route planning requests of the trip object can be obtained.

[0097] In one embodiment, for the obtained historical route planning requests 1-6 of the trip object, the historical real route planning requests of the trip object in the historical route planning requests 1-6 can also be determined in the following way:

[0098] Firstly, the historical route planning requests 1-6 are grouped into 5 sets of adjacent historical route planning requests, i.e., the historical route planning requests 1 and 2, 2 and 3, 3 and 4, 4 and 5, and 5 and 6 are respectively grouped into a set of adjacent historical route planning requests;

[0099] Then, for each set of adjacent historical route planning requests, whether the historical route planning requests 1-5 are historical real route planning requests of the trip object is determined by the steps S3021-S3026, if yes, the label of the corresponding historical route planning request is marked as 1, if not, the label of the corresponding historical route planning request is marked as 0;

[0100] Finally, the historical real route planning requests of the trip object are obtained by screening out the historical route planning requests corresponding to the label 1 in the historical route planning requests 1-5.

[0101] In the embodiments of the present application, for each set of two adjacent historical route planning requests, the second travel time interval of the two adjacent historical route planning requests can be calculated first. If the second travel time interval is less than or equal to a preset time threshold, the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests can be recorded as a non-historical real route planning request of the travel object. If the second travel time interval is greater than the preset time threshold, it is further determined whether the two adjacent historical route planning requests are same-trip en route planning requests. If the two historical route planning requests are same-trip en route planning requests, the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests can be recorded as a historical real route planning request of the travel object. If the two adjacent historical route planning requests are not same-trip en route planning requests, it is further determined whether the completion speed of the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests is less than a preset speed threshold. If yes, the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests is recorded as a historical real route planning request of the travel object. If no, the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests is recorded as a non-historical real route planning request of the travel object.

[0102] In the embodiments of the present application, when it is determined whether the historical route planning request with earlier historical travel time in the two adjacent historical route planning requests is a historical real route planning request of the travel object and is marked, the label of the finally determined historical real route planning request of the travel object can be recorded as 1, and the label of the finally determined non-historical real route planning request of the travel object can be recorded as 0. In this way, after all historical route planning requests are marked, the historical route planning requests with label 1 can be screened out to obtain all historical real route planning requests of the travel object.

[0103] The size of the preset time threshold used for comparison with the second travel time interval can be determined according to actual conditions or artificial experience, for example, it can be 1 minute or 5 minutes.

[0104] The preset speed threshold corresponds to the historical travel mode in the historical route planning data corresponding to the historical route planning request with early historical travel time, and specifically, different types of historical travel modes correspond to different preset speed thresholds, which can be determined according to actual conditions or artificial experience. For example, when the travel mode is driving, the preset speed threshold in the driving mode can be less than 200 km / h, when the travel mode is riding, the preset speed threshold in the riding mode can be less than 60 km / h, when the travel mode is taking public transportation, the preset speed threshold in the public transportation mode can be less than 150 km / h, and when the travel mode is walking, the preset speed threshold in the walking mode can be less than 30 km / h.

[0105] Embodiment two

[0106] The embodiment of the present application provides a training method of a travel mode prediction model, and the flow thereof is shown in Figure 2 The method comprises the following steps:

[0107] S201, obtaining route planning data corresponding to a historical route planning request as a sample as sample route planning data, the sample route planning data at least comprising: object starting point, object ending point, object travel mode, object travel time and object starting and ending point distance;

[0108] S202, obtaining route planning data corresponding to a historical route planning request with historical travel time earlier than the object travel time in the sample route planning data as historical route planning data, the historical route planning data at least comprising: historical travel mode, historical starting and ending point distance and historical travel time;

[0109] S203, obtaining feature data for model training based on the sample route planning data and the historical route planning data;

[0110] S204, taking the feature data as the input of the travel mode prediction model to be trained, training the model until the parameters of the model meet the requirements according to the object travel mode in the sample route planning data.

[0111] In the embodiment of the present application, the historical route planning request as a sample in the above step S201 is obtained by the following way:

[0112] S401: sorting the historical route planning requests of the travel object in the order from early to late according to historical travel time;

[0113] S402: determining the historical real route planning request of the travel object based on the historical travel time, historical travel mode, historical ending point and historical starting and ending point distance in the historical route planning data corresponding to the adjacent two historical route planning requests.

[0114] S403: Select a plurality of historical route planning requests as sample historical route planning requests from the obtained plurality of historical real route planning requests of the travel object.

[0115] In one embodiment, the implementation process of steps S401 and S402 described above is similar to the implementation process of steps S301 and S302 described in Embodiment I, and the detailed description can be referred to Embodiment I. In this embodiment, the detailed description is not repeated here.

[0116] In step S201, the object start-end point distance in the sample route planning data can be the straight-line distance between the object start point and the object end point. The object start point and the object end point can be the latitude and longitude position of the object start point and the latitude and longitude position of the object end point. The latitude and longitude position of the object start point can be obtained by positioning, and the positioning method can adopt the method in the prior art, such as GPS positioning, base station positioning, WIFI assisted positioning, etc. The latitude and longitude position of the object end point is generally determined by the end point selected by the travel object in the electronic map.

[0117] The object travel time can be the travel time selected by the travel object when initiating the historical real route planning request. For example, the travel object initiates a route planning request through a map navigation App at 15:00 in the afternoon, but the travel object selects a travel time of 15:20, so the travel time is 15:20 at this time. If the travel object does not select a travel time when initiating the route planning request, the time when the travel object initiates the route planning request can be used as the object travel time. For example, the travel object initiates a route planning request through a map navigation App at 15:00 in the afternoon, and the travel object does not select a travel time, so the travel time is 15:00 at this time.

[0118] At the same time, since the historical route planning request as sample is the historical real route planning request of the travel object, the object travel mode, i.e. the true value information of the travel mode of the travel object, can also be included in the sample route planning data.

[0119] In step S202, the historical route planning request whose historical travel time is earlier than the object travel time in the sample route planning data can be a historical real route planning request that can represent the real travel demand of the travel object, which is filtered from the historical route planning request of the travel object. In the embodiment of the present application, there can be some historical planning routes in the historical route planning request of the travel object that cannot represent the real travel demand of the travel object. This is because when performing route planning, the navigation server can recommend an incorrect travel mode, the travel object can also misoperate, or the travel object just wants to compare the results of the planned routes of various travel modes. Therefore, in order to obtain more accurate feature data, after determining the plurality of historical real route planning requests of the travel object through steps S401 and S402 and selecting a plurality of historical route planning requests as samples, the preset number of historical real route planning requests are selected from the plurality of historical real route planning requests of the travel object in the order from early to late according to the object travel time of the historical route planning request as a sample. When the preset number of historical real route planning requests are obtained, the route planning data corresponding to the preset number of historical real route planning requests are obtained as historical route planning data.

[0120] In step S203, the feature data for model training is obtained based on the sample route planning data and the historical route planning data. The feature data for model training can be a vector, and the object travel mode in the sample route planning data can be used as the travel mode true value for model training. Each information included in the feature data for model training can be represented by one or more parameters in the vector. For example, when the sample route planning data is object starting point, object ending point, object travel mode, object travel time, and object starting and ending point distance, since the object travel mode in the sample route planning data can be used as the travel mode true value for model training, it is not used as a feature parameter when training the model. If the object starting point and the object ending point are expressed by the latitude and longitude position, the object starting point is composed of the starting point longitude and the starting point latitude, and the object ending point is composed of the ending point longitude and the ending point latitude, so a 6-dimensional feature can be generated according to the sample route planning data. Assuming that the historical route planning data is the route planning data corresponding to n historical route planning requests, where n is a positive integer, and each historical route planning data is historical travel mode, historical starting and ending point distance, and historical travel time, a 3n-dimensional feature can be generated according to n historical route planning data. Finally, the feature data for model training obtained by combining the 6-dimensional feature and the 3n-dimensional feature is a vector including a 3n+6-dimensional feature.

[0121] Of course, in order to obtain richer information of the training feature data in the embodiment of the application, the above-mentioned sample route planning data can further include other information of the route planning request of the travel object, and the above-mentioned historical route planning data can further include other information of the historical route planning request, such as point of interest (POI) information, so that the obtained feature data can further include a POI parameter.

[0122] In the embodiment of the application, in order to make the obtained feature data for model training more convenient for operation and processing in the model training process, the parameter of the historical travel time can be replaced by a parameter of a travel time interval of the object travel time in the above-mentioned sample route planning data and the historical travel time in the historical route planning data, and then the above-mentioned step S203 can be implemented by the following steps:

[0123] determining a third travel time interval of the object travel time and the historical travel time based on the object travel time in the sample route planning data and the historical travel time in the historical route planning data;

[0124] generating feature data based on at least the third travel time interval, the historical travel mode, the historical start-end point distance, the object start point, the object end point, the object travel mode, the object travel time and the object start-end point distance.

[0125] The specific process of the step S204 described above until the parameter of the model meets the requirement according to the object travel mode in the sample route planning data includes the following steps:

[0126] comparing the object travel mode in the sample route planning data with the travel mode predicted by the travel mode prediction model, and the travel mode prediction model adjusts its own parameters according to the comparison result until the parameter of the model meets the requirement.

[0127] In the embodiment of the application, when training the travel mode prediction model, a training target can be set for the travel mode prediction model to be trained in advance, and the training target can be a loss function. The parameter of the model corresponding to the training target is the loss of the loss function, such as cross-entropy loss. In the model training process, by comparing the object travel mode in the sample route planning data of each historical route planning request as a sample with the travel mode predicted by the travel mode prediction model, the loss of the loss function is obtained, and the travel mode prediction model adjusts its own parameters according to the loss of the loss function to obtain an adjusted travel mode prediction model. The model training process is repeated until the loss of the loss function cannot be reduced, that is, the parameter of the model meets the requirement, and the model training is ended to obtain the trained travel mode prediction model.

[0128] In a preferred embodiment, during the model training process, due to the limited number of feature data during training, the travel mode prediction model obtained when the loss of the loss function is the smallest may be over-fitted, resulting in that the accuracy of the travel mode predicted by the travel mode prediction model may not be high. In order to solve the problem of over-fitting of the travel mode prediction model, in the embodiment of the present application, the feature data corresponding to a plurality of historical real route planning requests as samples can be divided into training samples and test samples, the training samples are used for model training, and during the training process of the travel mode prediction model, the test samples are repeatedly input into the travel mode prediction model to be trained, the object travel mode of the test sample is compared with the prediction result predicted by the travel mode prediction model, the accuracy of the travel mode prediction of the travel mode prediction model is obtained, when the accuracy of the travel mode prediction is the highest, it is determined that the parameters of the model meet the requirements, and the travel mode prediction model when the accuracy of the travel mode prediction is the highest is taken as the trained travel mode prediction model.

[0129] For example, the number of historical route planning requests as samples during training is 10000, and 10000 groups of feature data for model training are obtained. Then, 8000 groups of the 10000 groups of feature data for model training are taken as training samples, and the remaining 2000 groups are taken as test samples. When training the travel mode prediction model, the 8000 groups of feature data for model training as training samples are input into the travel mode prediction model to train the model, and during the training process, the 2000 groups of feature data for model training as test samples are repeatedly input into the travel mode prediction model being trained. Each input will obtain 2000 prediction results, and the 2000 prediction results are compared with the object travel mode true value of the historical real route planning request as sample, i.e. the object travel mode in the sample route planning data. If they are the same, the prediction result of the travel mode prediction model is correct; if they are different, the prediction result of the travel mode prediction model is incorrect. Finally, the number of correct prediction results among the 2000 prediction results is obtained, assuming that the number of correct prediction results is 1700, and the accuracy of the travel mode prediction model is 85%. If the number of times of inputting the 2000 test samples into the travel mode prediction model being trained during the training process of the travel mode prediction model is 10, the accuracy of the travel mode prediction model obtained by the 10 times of testing is: 40%, 70%, 72%, 77%, 81%, 83%, 84%, 85%, 84%, and 83%. Then, the travel mode prediction model with an accuracy of 85% can be selected as the trained travel mode prediction model, and the travel mode prediction model with the smallest loss of the loss function will not be selected as the trained travel mode prediction model.

[0130] In a preferred embodiment, if the accuracy of the travel mode prediction model obtained from the test sample is low each time after the model training ends, the travel mode prediction model obtained by training cannot meet the demand of travel mode recommendation, and the parameters of the travel mode prediction model can also be adjusted manually. The travel mode prediction model after parameter adjustment is trained to obtain the accuracy of the travel mode prediction model after parameter adjustment. The process of repeatedly determining the accuracy of the travel mode prediction model, adjusting the parameters according to the accuracy, and training the travel mode prediction model after parameter adjustment is repeated until the accuracy of the travel mode prediction model reaches the preset requirement, and a trained prediction model is obtained.

[0131] The inventors of the present application found in practice that if the method of directly using real-time route planning data of a travel object to predict travel mode is used, features are extracted from real-time route planning data of a travel object, for example, the object start-end distance, object start point, object end point, object travel time, etc. are used as feature parameters, and the travel mode requested by the travel object is used as the true value for model training to obtain a travel prediction model. Since this travel prediction model is trained, the true value of the travel mode used may be an incorrect travel mode recommended by the navigation server, or a travel mode misoperated by the travel object, or only the result of comparing various travel modes when the travel object plans the route. The route planning request in these scenarios is not the real travel demand of the travel object. Therefore, the real-time route planning data and the true value of the travel mode during training are not the real travel demand of the travel object, and the accuracy of the travel mode prediction model obtained by training cannot be guaranteed, resulting in that the predicted travel mode may not meet the real demand of the travel object.

[0132] Based on this, the inventors of the present application propose the training method of the travel mode prediction model as described above, by obtaining the historical real route planning request of the travel object, selecting the historical route planning request as a sample and using the route planning data corresponding to the historical route planning request as a sample as a sample route planning data, obtaining the route planning data corresponding to the historical route planning request with a historical travel time earlier than the object travel time in the sample route planning data as historical route planning data; based on the sample route planning data and the historical route planning data, the feature data for model training is obtained, the travel mode prediction model is trained using the feature data for model training, and when the preset model target parameter meets the requirement, a trained travel mode prediction model is obtained.

[0133] The feature data for model training is obtained by comprehensively using the sample route planning data of the travel object and a preset number of historical route planning data. The feature data for model training contains not only the real route planning data corresponding to the historical route planning request as a sample, but also the real historical route planning data corresponding to a preset number of historical real route planning requests whose travel time is earlier than the historical route planning request as a sample. Therefore, the accuracy of the travel mode prediction model obtained by training is higher, which is beneficial to better predict the travel mode. The travel mode predicted by using the trained travel mode prediction model is more in line with the real travel demand of the travel object, reducing the number of switching between travel modes when the travel object plans a route, thereby reducing the complexity of the travel object's route planning operation, saving the travel object's route planning time, improving the user experience of the travel object, and reducing the route calculation cost of the navigation server.

[0134] Embodiment Three

[0135] In one embodiment, referring to Figure 5 The travel mode recommendation method provided by the embodiment of the application can be executed by the navigation server 1 in the navigation service system. In the navigation service system, the navigation server 1 is connected with the terminal device 2 and the data server 3 through a network. When a travel object with a route planning demand initiates a route planning request through the terminal device 2, the terminal device 2 realizes network communication with the navigation server 1 through network access, and the navigation server 1 can obtain the route planning data of the travel object in real time in response to the route planning request of the terminal device 2. The route planning data of the travel object includes the object starting point, the object ending point, the object travel time and the object starting and ending point distance. At the same time, the navigation server 1 communicates with the data server 3 to obtain the data of the historical route planning request of the travel object and processes it to obtain the historical route planning data of the travel object. The historical route planning data of the travel object includes the historical travel mode, the historical starting and ending point distance and the historical travel time. The navigation server 1 generates feature data based on the route planning data of the travel object and the historical route planning data of the travel object, and inputs the feature data into a pre-trained travel mode prediction model to predict the travel mode recommended to the travel object. Therefore, the navigation server 1 can plan a route and make a recommendation based on the predicted travel mode recommended to the travel object, and finally send the determined travel mode and the corresponding recommended route to the terminal device 2. The travel mode and the route recommended to the travel object are more in line with the actual travel demand of the travel object.

[0136] The specific implementation mode of the navigation server 1 in the embodiment of the application for route planning based on the obtained travel mode of the travel object can refer to the description in the prior art, and the embodiment of the application does not limit it.

[0137] In the embodiments of the present application, the data of historical route planning requests in the data server 3 can be data obtained and uploaded by the navigation server 1 when planning a route for a travel object, or can be data uploaded by the terminal device after initiating a route planning request and obtaining a recommended planning route from the navigation server 1.

[0138] In the embodiments of the present application, the navigation server 1 and the data server 3 can be separately arranged, can be integrated into one server, or the navigation server 1 and the data server 3 can be implemented by a server cluster.

[0139] In the embodiments of the present application, the training process of the travel mode prediction model can also be performed in the navigation server 1. When training the travel mode prediction model, the navigation server 1 obtains and processes the data of historical route planning requests of travel objects by communicating with the data server 3, obtains historical route planning requests as samples, and further obtains sample route planning data corresponding to the historical route planning requests as samples and historical route planning data corresponding to historical route planning requests whose historical travel times are earlier than the object travel times in the sample route planning data, the sample route planning data including object starting points, object ending points, object travel modes, object travel times and object starting and ending point distances, and the historical route planning data including historical travel modes, historical starting and ending point distances and historical travel times. The navigation server 1 generates feature data for model training based on the sample route planning data and the historical route planning data, and inputs the travel mode prediction model to be trained for training, and finally obtains a trained travel mode prediction model.

[0140] In the embodiments of the present application, the navigation server 1 and the data server 3 can be separately arranged, can be integrated into one server, or the navigation server 1 and the data server 3 can be implemented by a server cluster.

[0141] The navigation service system provided by the embodiments of the present application can obtain recommended travel modes and recommended routes that meet the real route planning requirements after the terminal device 2 sends a route planning request to the navigation server 1, and provides better service and user experience for the navigation of the travel object.

[0142] In the navigation service system of this application embodiment, navigation server 1 generates feature data based on real-time acquired route planning data of the traveler and historical route planning data of the traveler obtained from data server 3. This feature data is then used in a pre-trained travel mode prediction model to predict and recommend travel modes to the traveler. During travel mode prediction, not only is real-time route planning information of the traveler acquired, but historical route planning information is also obtained from the traveler's historical route planning request data. By comprehensively utilizing both historical and real-time route planning information, the accuracy of travel mode prediction is improved. This not only enhances the user experience for the traveler but also significantly reduces the computational resource consumption of the navigation server.

[0143] The following example illustrates how a traveler uses a map navigation application on their terminal device to plan a route and obtain a recommended travel mode and route from the navigation server.

[0144] In one embodiment, refer to Figures 6a-6c As shown, assume the traveler runs a map navigation application on their terminal device. If the traveler launches the map navigation application, as follows: Figure 6a As shown, the map navigation application will display map information for the current location (i.e., "Beijing Capital International Airport"). When the traveler is in... Figure 6a When you enter your destination information (i.e., "Shoukai Plaza") in the search box of the map navigation application's interface and perform a destination search, the map navigation application sends a request to the navigation server to obtain the map data corresponding to the destination. After receiving the map data corresponding to the destination, the map navigation application's interface switches to the following... Figure 6b The search results interface shown will display map information of the destination location (i.e., "Shoukai Plaza") in the map navigation application interface. Assuming the traveler is in a location such as... Figure 6b Clicking the "Route" area in the search results interface sends a route planning request to the navigation server via the map navigation application. The navigation service then responds to this request by executing the travel mode recommendation method described in the above embodiments. This accurately predicts the travel mode needed by the traveler and performs route planning to obtain several recommended routes, which are then returned to the map navigation application for the traveler to choose from. Figure 6c The route planning recommendation interface shows that the recommended travel mode is "driving" mode, and displays the recommended route from the current location (i.e., "Beijing Capital International Airport") to the destination location (i.e., "Shoukai Plaza").

[0145] Based on the same inventive concept, the embodiments of the present application further provide a travel mode recommendation device, a travel mode prediction model training device, a travel mode recommendation system, a travel mode recommendation service, a model training service, a computer readable storage medium and a navigation device. Since the principles of the problems solved by these devices, terminals and systems are similar to those of the aforementioned travel mode recommendation method and travel mode prediction model training method, the implementation of the devices, equipment and systems can be referred to the implementation of the aforementioned method, and the repeated parts will not be described here.

[0146] The embodiments of the present application provide a travel mode recommendation device, as shown in Figure 7 , which comprises:

[0147] The first information acquisition module 101 is configured to obtain route planning data of a travel object in response to a route planning request, wherein the route planning data at least comprises a starting point, an ending point, a current request time and a starting- ending point distance;

[0148] The second information acquisition module 102 is configured to obtain historical route planning data of the travel object, wherein the historical route planning data at least comprises a historical travel mode, a historical starting- ending point distance and a historical request time;

[0149] The learning feature determination module 103 is configured to obtain feature data for model learning based on the historical route planning data and the route planning data;

[0150] The prediction module 104 is configured to input the feature data into a pre-trained travel mode prediction model to predict a travel mode recommended to the travel object.

[0151] In one embodiment, the learning feature determination module 103 is specifically configured to determine a first travel time interval between the object travel time and the historical travel time;

[0152] The feature data is generated based at least on the first travel time interval, the historical travel mode, the historical starting- ending point distance, the object starting point, the object ending point, the object travel time and the object starting- ending point distance.

[0153] In one embodiment, the second information acquisition module 102 is specifically configured to select a preset number of historical real route planning requests in order from early to late to the object travel time from the pre-screened historical real route planning requests of the travel object;

[0154] The historical route planning data corresponding to the preset number of historical real route planning requests is obtained.

[0155] In an embodiment, the second information obtaining module 102 is specifically configured to sort the historical route planning requests of the travel object in chronological order of historical travel times from early to late.

[0156] Based on the historical travel time, the historical travel mode, the historical end point and the historical start-end point distance in the historical route planning data corresponding to the two adjacent historical route planning requests, the historical real route planning request of the travel object is determined.

[0157] In an embodiment, the second information obtaining module 102 is specifically configured to obtain a second travel time interval of the two adjacent historical route planning requests based on the historical travel time in the historical route planning data corresponding to the two adjacent historical route planning requests.

[0158] Based on the historical end point, the historical start-end point distance and the historical travel mode in the historical route planning data corresponding to the two adjacent historical route planning requests, it is determined whether the two adjacent historical route planning requests are same-trip planning requests.

[0159] When the second travel time interval of the two adjacent historical route planning requests is greater than a preset time threshold and the two adjacent historical route planning requests are same-trip planning requests, the historical route planning request with an earlier historical travel time in the two adjacent historical route planning requests is recorded as the historical real route planning request of the travel object.

[0160] In an embodiment, the second information obtaining module 102 is specifically configured to obtain a request completion speed of the historical route planning request with an earlier historical travel time in the two adjacent historical route planning requests.

[0161] It is determined whether the request completion speed is less than a preset speed threshold, and the preset speed threshold is related to the historical travel mode in the historical route planning data corresponding to the historical route planning request with an earlier historical travel time.

[0162] When the second travel time interval of the two adjacent historical route planning requests is greater than a preset time threshold, the two adjacent historical route planning requests are not same-trip planning requests, and the request completion speed is less than the preset speed threshold, the historical route planning request with an earlier historical travel time in the two adjacent historical route planning requests is recorded as the historical real route planning request of the travel object.

[0163] In an embodiment, the second information obtaining module 102 is specifically configured to determine the request completion speed of the historical route planning request with an earlier historical travel time based on the historical start-end point distance in the historical route planning data corresponding to the historical route planning request with an earlier historical travel time and the second travel time interval.

[0164] In one embodiment, the second information obtaining module 102 is specifically configured to determine whether the historical end point and the historical travel mode in the historical route planning data corresponding to the two adjacent historical route planning requests are the same, and determine whether the historical start-end point distance in the historical route planning data corresponding to the historical route planning request with the later historical travel time is less than the historical start-end point distance in the historical route planning data corresponding to the historical route planning request with the earlier historical travel time.

[0165] If yes, it is determined that the historical route planning request with the later historical travel time is a same-trip planning request of the historical route planning request with the earlier historical travel time.

[0166] The embodiments of the present application provide a training device of a travel mode prediction model, as shown in Figure 8 The training device comprises:

[0167] A third information obtaining module 201 is configured to obtain route planning data corresponding to a historical route planning request as a sample as sample route planning data, and the route planning data of the sample route planning request at least comprises an object start point, an object end point, an object travel mode, an object travel time and an object start-end point distance.

[0168] A fourth information obtaining module 202 is configured to obtain route planning data corresponding to a historical route planning request with an earlier historical travel time than the object travel time in the sample route planning data as historical route planning data, and the historical route planning data at least comprises a historical planning travel mode, a historical planning travel distance and a historical planning travel time.

[0169] A training feature determining module 203 is configured to obtain feature data for model training based on the sample route planning data and the historical route planning data.

[0170] A model training module 204 is configured to take the feature data as an input of a travel mode prediction model to be trained, train the model until the parameters of the model meet the requirements according to the object travel mode in the sample route planning data.

[0171] In one embodiment, the training feature determining module 203 is specifically configured to determine a third travel time interval between the object travel time in the sample route planning data and the historical travel time in the historical route planning data.

[0172] Generate feature data based on at least the third travel time interval, the historical travel mode, the historical start-end point distance, the object start point, the object end point, the object travel mode, the object travel time, and the object start-end point distance.

[0173] In one embodiment, the model training module 204 described above is specifically configured to compare the object travel mode in the sample route planning data with the travel mode predicted by the travel mode prediction model, and the travel mode prediction model adjusts its parameters according to the comparison result until the parameters of the model meet the requirements.

[0174] In one embodiment, the training device of the travel mode prediction model described above further includes a sample acquisition module 200 configured to select a plurality of historical route planning requests as samples from historical route planning requests of travel objects.

[0175] In one embodiment, the sample acquisition module 200 described above is specifically configured to sort the historical route planning requests of the travel objects in the order from early to late historical travel time;

[0176] Determine the historical real route planning request of the travel object based on the historical travel time, the historical travel mode, the historical end point, and the historical start-end point distance in the historical route planning data corresponding to the two adjacent historical route planning requests.

[0177] Select a plurality of historical route planning requests as samples from the obtained plurality of historical real route planning requests of the travel object.

[0178] In one embodiment, the sample acquisition module 200 described above is specifically configured to select a preset number of historical real route planning requests from the obtained plurality of historical real route planning requests of the travel object in the order from early to late object travel time in the sample route planning data;

[0179] The fourth information acquisition module 202 described above is specifically configured to acquire the route planning data corresponding to the preset number of historical real route planning requests as historical route planning data.

[0180] Embodiments of the present application provide a travel mode recommendation system, as shown in Figure 9 The navigation server 1 and at least one terminal device 2 are included, wherein:

[0181] The travel mode recommendation device described above is arranged in the navigation server 1, configured to receive the route planning request sent by the terminal device 2, and send the determined travel mode to the terminal device 2.

[0182] The terminal device 2 is configured to send the route planning request to the navigation server 1 and receive the travel mode returned by the navigation server 1.

[0183] The embodiments of the present application provide a travel mode recommendation service, which performs the travel mode recommendation method described above when running.

[0184] The embodiments of the present application provide a model training service, which performs the training method of the travel mode prediction model described above when running.

[0185] The embodiments of the present application provide a computer readable storage medium, which stores computer instructions, and the instructions are executed by a processor to implement the travel mode recommendation method described above and / or the training method of the travel mode prediction model described above.

[0186] The embodiments of the present application provide a navigation device, which includes a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the travel mode recommendation method described above and / or the training method of the travel mode prediction model described above when executing the program.

[0187] The embodiments of the present application provide a travel mode recommendation method, related method and device and system, in the travel mode recommendation method, route planning data and historical route planning data of a travel object are obtained, feature data is obtained based on the route planning data and the historical route planning data, the travel mode recommended to the travel object is predicted by using the feature data through a pre-trained travel mode prediction model, since the feature data integrates the historical data (historical route planning data) and real-time data (route planning data in the route planning request) of the travel object, and the travel mode prediction model better predicts the travel mode, the accuracy of the travel mode prediction is improved, on one hand, the number of switching between travel modes when the travel object plans a route is reduced, thereby reducing the operation complexity of the travel object in route planning, saving the route planning time of the travel object, improving the user experience of the travel object, and meanwhile, the route calculation cost of the navigation server is reduced.

[0188] Unless specifically stated otherwise, terms such as processing, computing, calculating, determining, displaying, and the like, can refer to an action or process of one or more processing or computing systems, or similar devices, that manipulate or transform data represented as physical (e.g., electronic) quantities within the systems' registers or memories into other data similarly represented as physical quantities within the systems' memories, registers or other such information storage, transmission or display devices. The terms "information," "data," "instructions," “command,” “signal,” “bit,” “symbol,” and “chip” refer to physical quantities that can be measured, processed, transformed, or otherwise manipulated by a processing system.

[0189] It should be understood that the specific order or hierarchy of steps in the processes disclosed is an example of exemplary approaches. Based upon design preferences, it should be understood that the specific order or hierarchy of steps in the processes can be re-arranged while remaining within the scope of the application. The accompanying method claims present elements of the various steps in a sample order, and are not meant to be limited to the specific order or hierarchy presented.

[0190] In the above detailed description, various features are grouped together in single embodiments for the purpose of streamlining the disclosure. This disclosed approach is not to be interpreted as reflecting a necessity that the claimed application require more features than are explicitly recited in each claim. On the contrary, as indicated, the application is directed to each individual feature individually recited in each claim, and anything that takes those features in combination, provided it falls within the definition of the term used. Thus, the claims following, are hereby expressly incorporated into this detailed description, where each claim by itself is made a separate preferred embodiment of the application.

[0191] Those skilled in the art will further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans can implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

[0192] The steps of a method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium can be integral to the processor. The processor and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor and the storage medium can reside as discrete components in a user terminal.

[0193] For a software implementation, the techniques described herein can be implemented with modules (e.g., procedures, functions, and so on) that perform the functions described herein. The software codes can be stored in memory units and executed by processors. The memory unit can be implemented within the processor or external to the processor, in which case it can be communicatively coupled to the processor via various means as is known in the art.

[0194] The above description includes one or more examples of the embodiments. Of course, not all possible combinations of components or methods described above can be claimed as embodiments. One of ordinary skill in the art can recognize that modifications and variations of the described embodiments can be made without departing from the scope of the present disclosure. It is therefore intended that the embodiments described herein be considered in all respects as illustrative and not restrictive, particularly as numerous modifications and further embodiments can become apparent to those skilled in the art. Accordingly, the scope of the present disclosure is intended to be defined by the following claims rather than the description. Moreover, the use of the terms "first", "second", etc. do not denote any order or importance, but rather the terms are used to distinguish one element from another. Furthermore, the use of the terms "including", "containing", etc. are meant to encompass the terms "consisting of" and / or "consisting essentially of". Moreover, the use of the term "or" is meant to encompass "and / or", unless otherwise indicated.

Claims

1. A travel mode recommendation method, comprising: obtaining route planning data of a travel object in response to a route planning request, the route planning data at least including: object origin, object destination, object travel time and object origin-destination distance; obtaining historical route planning data of the travel object, the historical route planning data at least including: historical travel mode, historical origin-destination distance and historical travel time; generating feature data based on the historical route planning data and the route planning data; inputting the feature data into a pre-trained travel mode prediction model to predict a travel mode recommended to the travel object; wherein the obtaining of the historical route planning data of the travel object comprises: sorting historical route planning requests of the travel object in order of historical travel time from early to late; determining a historical real route planning request of the travel object based on historical travel time, historical travel mode, historical destination and historical origin-destination distance in historical route planning data corresponding to adjacent two historical route planning requests; selecting a preset number of historical real route planning requests in order of object travel time from early to late from the pre-screened historical real route planning requests of the travel object; obtaining historical route planning data corresponding to the preset number of historical real route planning requests.

2. The method of claim 1, wherein, The generating of the feature data based on the historical route planning data and the route planning data comprises: determining a first travel time interval between the object travel time and the historical travel time; generating feature data based on at least the first travel time interval, the historical travel mode, the historical origin-destination distance, the object origin, the object destination, the object travel time and the object origin-destination distance.

3. The method of claim 1, wherein, The determining of the historical real route planning request of the travel object based on historical travel time, historical travel mode, historical destination and historical origin-destination distance in historical route planning data corresponding to adjacent two historical route planning requests comprises: obtaining a second travel time interval of the adjacent two historical route planning requests based on historical travel time in historical route planning data corresponding to the adjacent two historical route planning requests; obtaining whether the adjacent two historical route planning requests are same-trip planning requests based on historical destination, historical origin-destination distance and historical travel mode in historical route planning data corresponding to the adjacent two historical route planning requests; when the second travel time interval of the adjacent two historical route planning requests is greater than a preset time threshold and the adjacent two historical route planning requests are same-trip planning requests, recording the historical route planning request with earlier historical travel time in the adjacent two historical route planning requests as the historical real route planning request of the travel object.

4. The method of claim 3, wherein, The method further comprises: obtaining a request completion speed of the historical route planning request with earlier historical travel time in the adjacent two historical route planning requests. It is determined whether the request completion speed is less than a preset speed threshold, which is related to the historical travel patterns in the historical route planning data corresponding to historical route planning requests with earlier historical travel times; If the second travel time interval between two adjacent historical route planning requests is greater than a preset time threshold, the two adjacent historical route planning requests are not related to the same-journey planning request, and the request completion speed is less than the preset speed threshold, then the historical route planning request with the earlier historical travel time among the two adjacent historical route planning requests is recorded as the historical true route planning request of the travel object.

5. The method of claim 4, wherein, The step of obtaining the request completion speed of the historical route planning request with the earlier historical travel time among two adjacent historical route planning requests specifically includes: Based on the historical origin-destination distance and the second travel time interval in the historical route planning data corresponding to the historical route planning request with the earlier historical travel time among the two adjacent historical route planning requests, the request completion speed of the historical route planning request with the earlier historical travel time is determined.

6. The method of claim 3, wherein, The step of determining whether two adjacent historical route planning requests are concurrent route planning requests based on the historical destination, historical origin-destination distance, and historical travel patterns in the historical route planning data corresponding to the two adjacent historical route planning requests specifically includes: Determine whether the historical destination and historical travel mode in the historical route planning data corresponding to the two adjacent historical route planning requests are the same, and determine whether the historical origin-destination distance in the historical route planning data corresponding to the historical route planning request with the later historical travel time is less than the historical origin-destination distance in the historical route planning data corresponding to the historical route planning request with the earlier historical travel time. If so, then the historical route planning request with the later historical travel time is determined to be the same route planning request as the historical route planning request with the earlier historical travel time.

7. A training method for a travel pattern prediction model, comprising: Sort the historical route planning requests of the travelers in order of their historical travel time from earliest to latest; Based on the historical travel time, historical travel mode, historical destination and historical origin-destination distance in the historical route planning data corresponding to two adjacent historical route planning requests, the historical true route planning request of the travel object is determined. Select multiple historical route planning requests as samples from the multiple historical real route planning requests obtained from the travel object; Obtain the route planning data corresponding to the historical route planning request used as a sample as sample route planning data. The sample route planning data includes at least: object origin, object destination, object travel mode, object travel time, and object origin-destination distance. The historical route planning data is obtained by acquiring the route planning data corresponding to the historical route planning request whose historical travel time is earlier than the travel time of the object in the sample route planning data. The historical route planning data includes at least: historical travel mode, historical origin-destination distance and historical travel time. Based on the sample route planning data and the historical route planning data, feature data for model training is obtained; The feature data is taken as input of a travel mode prediction model to be trained, and the model is trained until parameters of the model meet requirements according to an object travel mode in the sample route planning data.

8. The method of claim 7, wherein, The feature data for model training is obtained based on the sample route planning data and the historical route planning data, and includes: A third travel time interval of the object travel time and the historical travel time is determined based on the object travel time in the sample route planning data and the historical travel time in the historical route planning data; The feature data is generated based on at least the third travel time interval, the historical travel mode, the historical start-end point distance, the object start point, the object end point, the object travel mode, the object travel time and the object start-end point distance.

9. The method of claim 7, wherein, The parameters of the model meet requirements according to the object travel mode in the sample route planning data, and include: The object travel mode in the sample route planning data is compared with a travel mode predicted by the travel mode prediction model, and the travel mode prediction model adjusts its parameters according to a comparison result until the parameters of the model meet requirements.

10. The method of claim 7, wherein, The historical route planning data is obtained by selecting a preset number of historical real route planning requests from the obtained historical route planning requests of the travel object in a sequence from early to late according to the object travel time in the sample route planning data, and includes: A preset number of historical real route planning requests are selected from the obtained historical route planning requests of the travel object in a sequence from early to late according to the object travel time in the sample route planning data; Route planning data corresponding to the preset number of historical real route planning requests is obtained as the historical route planning data.

11. A travel mode recommendation device, comprising: A first information acquisition module is configured to obtain route planning data of a travel object in response to a route planning request, wherein the route planning data at least includes an object start point, an object end point, an object travel time and an object start-end point distance; A second information acquisition module is configured to obtain historical route planning data of the travel object, wherein the historical route planning data at least includes a historical travel mode, a historical start-end point distance and a historical travel time; A learning feature determination module is configured to obtain feature data for model learning based on the historical route planning data and the route planning data; A prediction module is configured to input the feature data into a pre-trained travel mode prediction model to predict a travel mode recommended to the travel object; The second information acquisition module is configured to: Sort historical route planning requests of the travel object in a sequence from early to late according to historical travel times; Determine historical real route planning requests of the travel object based on historical travel times, historical travel modes, historical end points and historical start-end point distances in historical route planning data corresponding to adjacent two historical route planning requests; Select a preset number of historical real route planning requests of the travel object in a sequence from early to late according to the object travel time from the pre-screened historical real route planning requests of the travel object; and acquire historical route planning data corresponding to the preset number of historical real route planning requests.

12. A device for training a travel mode prediction model, comprising: a sample acquisition module configured to sort historical route planning requests of a travel object by historical travel time from early to late; determine historical real route planning requests of the travel object based on historical travel time, historical travel mode, historical end point, and historical start-end point distance in historical route planning data corresponding to adjacent two historical route planning requests; select a plurality of historical route planning requests as samples from the obtained historical real route planning requests of the travel object; a third information acquisition module configured to acquire route planning data corresponding to the historical route planning requests as samples as sample route planning data, the sample route planning data at least including object start point, object end point, object travel mode, object travel time, and object start-end point distance; a fourth information acquisition module configured to acquire route planning data corresponding to historical route planning requests with historical travel time earlier than the object travel time in the sample route planning data as historical route planning data, the historical route planning data at least including historical travel mode, historical start-end point distance, and historical travel time; a training feature determination module configured to obtain feature data for model training based on the sample route planning data and the historical route planning data; a model training module configured to train the model by taking the feature data as input of the travel mode prediction model to be trained until the parameters of the model meet the requirements according to the object travel mode in the sample route planning data.

13. A travel mode recommendation system, comprising: a navigation server and at least one terminal device, wherein: the navigation server is provided with the travel mode recommendation device of claim 11, configured to receive route planning requests sent by the terminal device, and send the determined travel mode to the terminal device; the terminal device is configured to send the route planning request to the navigation server, and receive the travel mode returned by the navigation server.

14. A travel mode recommendation server, the server executes the travel mode recommendation method of any one of claims 1-6 when running.

15. A model training server, the server executes the travel mode prediction model training method of any one of claims 7-10 when running.

Images