Travel time prediction methods, model training methods, electronic devices and software products

Abstract

The travel time estimation method, model training method, electronic equipment, and program products disclosed herein relate to autonomous driving technology, intelligent transportation technology, vehicle-to-everything (V2X) technology, and intelligent cockpit technology. The method includes obtaining a target travel route corresponding to a target route planning request; obtaining information on other travel routes and determining, based on the information on other routes, the first road condition characteristics of the target road segment and the second road condition characteristics of other road segments connected to the target road segment; and determining the total estimated travel time of the target travel route based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment. The solution provided in this disclosure can estimate the time required to pass through a target road segment based on predicted road condition information, enabling a more accurate prediction of the time required to pass through the target road segment in the future, thereby obtaining a more accurate total estimated travel time required to pass through the target route.

Description

Technical Field

[0001] This disclosure relates to autonomous driving technology, intelligent transportation technology, vehicle networking technology, and intelligent cockpit technology in the field of artificial intelligence, and particularly to a method for predicting travel time, a model training method, electronic equipment, and software products. Background Technology

[0002] Before setting off on a road trip, users can use navigation software to plan routes and view the time required to reach their destination. For example, users can input their origin and destination into the navigation software, which can then plan a route from the origin to the destination and estimate the time required for each route.

[0003] Currently, navigation software determines the time required to reach the destination based on the speed limit data of each road included in the planned route, or estimates the time required to reach the destination by combining the user's vehicle's historical speed.

[0004] However, these methods are all based on static data to estimate the time required for a vehicle to reach its destination. However, future road conditions are dynamic and will affect this time. Therefore, the existing technology solutions cannot accurately estimate the time required to reach the destination by planning the route, resulting in a poor user experience. Summary of the Invention

[0005] This disclosure provides a method for predicting travel time, a model training method, an electronic device, and a program product for accurately predicting the travel time required to travel through a planned route.

[0006] According to a first aspect of this disclosure, a method for estimating travel time is provided, comprising:

[0007] Obtain the target route corresponding to the target route planning request, wherein the target route includes multiple target road segments along the route;

[0008] Obtain information on other routes corresponding to other route planning requests, and determine the first traffic condition characteristics of the target road segment and the second traffic condition characteristics of other road segments connected to the target road segment based on the information of the other routes;

[0009] Based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment, the first estimated travel time of the target road segment is determined;

[0010] The total estimated travel time of the target route is determined based on the first estimated travel time of each target road segment.

[0011] According to a second aspect of this disclosure, a model training method for estimating the travel time of a road segment is provided, comprising:

[0012] Obtain a training dataset, which includes a first traffic condition feature of the target road segment, a second traffic condition feature of other road segments connected to the target road segment, and a first travel time corresponding to the target road segment; wherein the first traffic condition feature and the second traffic condition feature are determined based on historical planned routes;

[0013] The first road condition feature and the second road condition feature are input into a preset model to obtain the predicted travel time corresponding to the target road segment;

[0014] The parameters in the model are adjusted based on the first travel time and the predicted travel time to obtain a model for estimating the travel time of a road segment.

[0015] According to a third aspect of this disclosure, a travel time estimation device is provided, comprising:

[0016] The route acquisition unit is used to acquire the target route corresponding to the target route planning request, wherein the target route includes multiple target road segments.

[0017] The route acquisition unit is also used to acquire information on other routes corresponding to other route planning requests;

[0018] The feature extraction unit is used to determine the first road condition features of the target road segment and the second road condition features of other road segments connected to the target road segment based on the information of the other travel routes.

[0019] The road segment travel time determination unit is used to determine the first estimated travel time of the target road segment based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment;

[0020] The route travel time determination unit is used to determine the total estimated travel time of the target travel route based on the first estimated travel time of each target road segment.

[0021] According to a fourth aspect of this disclosure, a model training apparatus for estimating the travel time of a road segment is provided, comprising:

[0022] The training data acquisition unit is used to acquire a training dataset, which includes a first traffic condition feature of the target road segment, a second traffic condition feature of other road segments connected to the target road segment, and a first travel time corresponding to the target road segment; wherein the first traffic condition feature and the second traffic condition feature are determined based on historical planned routes.

[0023] The prediction unit is used to input the first road condition feature and the second road condition feature into a preset model to obtain the predicted travel time corresponding to the target road segment;

[0024] An adjustment unit is used to adjust the parameters in the model according to the first travel time and the predicted travel time to obtain a model for estimating the travel time of a road segment.

[0025] According to a fifth aspect of this disclosure, an electronic device is provided, comprising:

[0026] At least one processor; and

[0027] A memory communicatively connected to the at least one processor; wherein,

[0028] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method as described in the first or second aspect.

[0029] According to a sixth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method as described in the first or second aspect.

[0030] According to a seventh aspect of this disclosure, a computer program product is provided, comprising: a computer program stored in a readable storage medium, at least one processor of an electronic device being able to read the computer program from the readable storage medium, and the at least one processor executing the computer program causing the electronic device to perform the method of the first aspect or the second aspect.

[0031] The travel time estimation method, model training method, electronic device, and program product disclosed herein include: obtaining a target travel route corresponding to a target route planning request, the target travel route including multiple target road segments; obtaining information on other travel routes corresponding to other route planning requests, and determining a first road condition feature of the target road segment and a second road condition feature of other road segments connected to the target road segment based on the information of the other routes; determining a first estimated travel time of the target road segment based on the first road condition feature of the target road segment and the second road condition feature corresponding to the target road segment; and determining the total estimated travel time of the target route based on the first estimated travel time of each target road segment. The solution provided in this disclosure can predict the future road conditions of each target road segment based on information from other existing routes, and then predict the first estimated travel time required to travel through the target road segment based on the road conditions of each target road segment. This method of estimating the time to travel through the target road segment based on predicted road condition information can more accurately predict the time required to travel through the target road segment in the future, and thus obtain a more accurate total estimated travel time required to travel through the target route.

[0032] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0033] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0034] Figure 1 A user terminal interface diagram illustrating an exemplary embodiment;

[0035] Figure 2 A flowchart illustrating a method for estimating travel time, as shown in an exemplary embodiment of this disclosure;

[0036] Figure 3 A flowchart illustrating a method for estimating travel time, as shown in another exemplary embodiment of this disclosure;

[0037] Figure 4 This is a schematic diagram illustrating a process for determining the travel time of each target road segment, as shown in an exemplary embodiment of this disclosure.

[0038] Figure 5 This is a schematic flowchart illustrating a model training method for estimating the travel time of a road segment, as shown in an exemplary embodiment of this disclosure.

[0039] Figure 6 This is a flowchart illustrating a model training method for estimating the travel time of a road segment, as shown in another exemplary embodiment of this disclosure.

[0040] Figure 7 A schematic diagram of the structure of a travel time estimation device shown in an exemplary embodiment of the present disclosure;

[0041] Figure 8 A schematic diagram of the structure of a travel time estimation device shown in another exemplary embodiment of this disclosure;

[0042] Figure 9 This is a schematic diagram of the structure of a model training device for estimating the travel time of a road segment, as shown in an exemplary embodiment of the present disclosure.

[0043] Figure 10 This is a schematic diagram of the structure of a model training device for estimating the travel time of a road segment, as shown in another exemplary embodiment of the present disclosure.

[0044] Figure 11 This is a block diagram of an electronic device used to implement the methods of the embodiments of this disclosure. Detailed Implementation

[0045] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0046] Figure 1 This is an example of a user terminal interface diagram.

[0047] like Figure 1 As shown, the user can input the departure point 11 and the destination 12 in the terminal. The terminal device can send a route planning request to the server. The request includes the departure point information and the destination information. The server can plan a reasonable route based on these two pieces of information, estimate the travel time of each route, and provide feedback to the terminal device with a better route based on the travel time.

[0048] After receiving the travel routes provided by the user, the terminal device can display these routes, such as... Figure 1 As shown in 13, the terminal device can also display the estimated travel time for each route.

[0049] Users typically refer to the estimated travel time of routes before choosing a route to take. Therefore, the accuracy of the estimated travel time has a significant impact on user experience.

[0050] If the estimated travel time for a route is based solely on the user's historical driving speed and the speed limit data for each section of the route, the estimated travel time may be inaccurate due to real-time changes in road conditions.

[0051] To solve the above-mentioned technical problems, the solution provided in this disclosure can, after planning the target route, use the route planning requests of multiple users to predict the first estimated travel time of each target road segment included in the target route, thereby obtaining the total estimated travel time of the entire target route.

[0052] Figure 2 This is a flowchart illustrating a method for estimating travel time as an exemplary embodiment of the present disclosure.

[0053] like Figure 2 As shown, the method for estimating travel time provided in this disclosure includes:

[0054] Step 201: Obtain the target route corresponding to the target route planning request. The target route includes multiple target road segments.

[0055] The solution provided in this disclosure can be applied to the backend server of navigation software.

[0056] In several scenarios, the server can receive target route planning requests. For example, a user can operate a user terminal, inputting the origin and destination, and clicking the search button, causing the user terminal to send a target route planning request to the server, including the origin and destination. Another example is when the user terminal's location deviates from the planned route, the server can receive the target route planning request from the user terminal. Yet another example is when the user terminal's location updates, which can also trigger the server to obtain a new target route planning request.

[0057] Specifically, the server can plan a target route from the origin to the destination based on the target route planning request. For example, from location A to location B, the server can plan multiple target routes. For instance, it can first enter target route L1 from location A, then pass through target route L2, and then through target route L3 to reach location B.

[0058] Furthermore, for each target travel route, the server can determine its corresponding estimated travel time based on the method provided in this disclosure.

[0059] Step 202: Obtain information on other routes corresponding to other route planning requests, and determine the first road condition characteristics of the target road segment and the second road condition characteristics of other road segments connected to the target road segment based on the information of other routes.

[0060] In practical applications, before estimating the travel time of the target route, alternative routes are usually determined based on other route planning requests. For example, if the estimated travel time of the target route is at time T, before time T, another user terminal may send an alternative route planning request to the server, and the server can plan alternative routes for it based on this request.

[0061] Typically, a server will plan multiple routes for an alternative route planning request. The server can then select the route that the user is most likely to choose as the alternative route. For example, the server can estimate the travel time for each route in the alternative route planning request based on the method provided in this disclosure, and then select an alternative route based on that travel time.

[0062] Among them, by using the other routes determined by the server for each other route planning request, the traffic conditions of each road segment in each future time segment can be predicted, and then the travel time of the target route can be predicted based on the traffic conditions of each target road segment included in the target route.

[0063] Specifically, the target travel route includes multiple target road segments. For each target road segment, the server can determine the first road condition feature based on information from other travel routes. The first road condition feature is used to characterize the road condition information of the target road segment extracted when predicting the travel time of the target travel route. For example, the number of vehicles that may arrive on the target road segment in each time period, and the average speed of vehicles on the target road segment.

[0064] Furthermore, other road segments may connect to the target road segment, and the road conditions of these other road segments typically affect the road conditions of the target road segment. Therefore, the method provided in this disclosure can also determine second road condition characteristics of other road segments connected to the target road segment based on information from other travel routes. These second road condition characteristics are used to characterize the road condition information of other road segments when predicting the travel time of the target travel route.

[0065] In practical applications, the method for obtaining the second road condition features of other road segments can be the same as the method for obtaining the first road condition features of the target road segment.

[0066] The information regarding other travel routes may include multiple road segments that need to be traversed, as well as predicted arrival times for each segment. For example, the information for an other travel route may include segment L4 and the arrival time t4, segment L5 and the arrival time t5, and segment L6 and the arrival time t6. The arrival time for any of the road segments may be determined based on the scheme disclosed herein, or it may be roughly estimated based on existing schemes in the prior art.

[0067] Other routes can be cached on the server, and these can be other routes currently in use on the user's terminal. For example, after the user's terminal issues a navigation request, the server can generate the target route based on this request and send the target route back to the user's terminal.

[0068] The server can also cache other routes. If a user stops the navigation function or exits the navigation software, the server can delete the corresponding other routes. This implementation method can obtain more routes that the user may drive on, thereby more accurately predicting future traffic conditions based on these target routes.

[0069] Step 203: Determine the first estimated travel time of the target road segment based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment.

[0070] The first road condition feature represents the road conditions of the target road segment in the future, while the second road condition feature represents the road conditions of other road segments that connect to the target road segment in the future. Therefore, by combining these features, the travel time for each target road segment can be estimated.

[0071] For example, if it is predicted that the number of vehicles on the target road segment will increase significantly in the next 5 minutes, and the number of vehicles on other road segments connected to the target road segment will also increase significantly, then the travel time on the target road segment can be estimated to increase in 5 minutes. Conversely, if it is predicted that the number of vehicles on the target road segment will decrease significantly in the next 30 minutes, and the number of vehicles on other road segments connected to the target road segment will also decrease significantly, then the travel time on the target road segment can be estimated to decrease in 30 minutes.

[0072] In one optional implementation, a time prediction model can be pre-set, which can be pre-trained. Specifically, it can be trained using training data to obtain the time prediction model. For example, route planning requests pre-cached before time t can be used to extract the features of road segments and the features of other road segments adjacent to the road segment. Historical logs can also be used to determine the actual travel time of the road segment in each time segment after time t. This actual travel time can then be used as supervision data, and the features of the road segment and the features of other road segments adjacent to the road segment can be used as input data to train the model. This allows the model to learn the ability to calculate the travel time of the road segment in each future time segment using the input traffic features.

[0073] Furthermore, after training the time prediction model, it can be deployed on a server. The target travel route includes multiple target road segments. The server can input the first traffic condition feature of a target road segment and the second traffic condition feature of other road segments connected to the target road segment into the deployed time prediction model to obtain the first estimated travel time of the target road segment.

[0074] Step 205: Determine the total estimated travel time of the route based on the first estimated travel time of each target road segment.

[0075] After obtaining the first estimated travel time of each target segment in the target travel route, the server can generate the total estimated travel time of the target travel route based on the first estimated travel time of each target segment.

[0076] Specifically, the server can overlay the first estimated travel time of each target road segment to obtain the total estimated travel time of the target route.

[0077] If the server can obtain the travel duration of each target road segment in each time segment, for any target road segment, the server can determine the time to arrive at the target road segment, and then determine the travel duration corresponding to the time segment to which that time belongs as the time required to travel the target road segment. Finally, by adding up the travel duration of each target road segment, the total estimated travel time of the target route can be obtained.

[0078] Furthermore, for example, if the target travel route includes target road segments L1, L2, and L3, it is possible to predict the travel duration of L1 within multiple time segments, the travel duration of L2 within multiple time segments, and the travel duration of L3 within multiple time segments.

[0079] In practical applications, based on the departure time t0, the travel duration T1 of the time segment to which t0 belongs can be determined from the travel durations of each travel segment in the target road segment L1. Then, (t0+T1) can be considered as the arrival time in L2. Next, the travel duration T2 of the time segment to which (t0+T1) belongs can be determined from the travel durations of each travel segment in the target road segment L2. Then, (t0+T1+T2) can be considered as the arrival time in L3. Finally, the travel duration T3 of the time segment to which (t0+T1+T2) belongs can be determined from the travel durations of each travel segment in the target road segment L3.

[0080] Among them, (T1+T2+T3) can be determined as the total estimated travel time of the target route, and (t0+T1+T2+T3) can be determined as the estimated arrival time when traveling along the target route.

[0081] The method for estimating travel time provided in this disclosure includes: obtaining a target route corresponding to a target route planning request, the target route including multiple target road segments; obtaining information on other routes corresponding to other route planning requests, and determining a first road condition characteristic of the target road segment and a second road condition characteristic of other road segments connected to the target road segment based on the information of other routes; determining a first estimated travel time of the target road segment based on the first road condition characteristic of the target road segment and the second road condition characteristic corresponding to the target road segment; and determining the total estimated travel time of the target route based on the first estimated travel time of each target road segment. In the solution provided in this disclosure, the future road conditions of each target road segment can be predicted based on existing information from other routes, and thus the first estimated travel time required to travel through that target road segment can be predicted based on the road conditions of each target road segment. This method of estimating the time required to travel through a target road segment based on predicted road condition information can more accurately predict the time required to travel through the target road segment in the future, thereby obtaining a more accurate total estimated travel time required to travel through the target route.

[0082] Figure 3 This is a flowchart illustrating a method for estimating travel time, which is another exemplary embodiment of this disclosure.

[0083] like Figure 3 As shown, the method for estimating travel time provided in this disclosure includes:

[0084] Step 301: Obtain the target route corresponding to the target route planning request. The target route includes multiple target road segments.

[0085] The implementation of step 301 is similar to that of step 201, and will not be described again.

[0086] Step 302: Obtain information on other routes corresponding to other route planning requests.

[0087] Step 302 is similar to the corresponding content shown in step 202, and will not be repeated here.

[0088] Step 303: Based on information from other travel routes, determine the first time distribution information of the number of vehicles arriving at the target road segment in the future, and the second time distribution information of the number of vehicles arriving at other road segments in the future.

[0089] The information regarding other travel routes includes the road segments traversed, and may also include the time taken to reach each road segment when traveling along other routes. For example, the time to reach each road segment can be determined based on the method of this disclosure, or it can be determined based on existing technology.

[0090] Specifically, by obtaining the arrival times of each road segment from information on other travel routes, we can obtain the first time distribution information of the target road segment, as well as the second time distribution information of other road segments connected to the target road segment.

[0091] Temporal distribution information is used to characterize the number of vehicles arriving at a road segment at different times, that is, to determine the distribution of vehicles arriving at the road segment according to the time dimension. By using the first and second temporal distribution information, we can obtain the number of vehicles on the target road segment and other road segments at different future times. Therefore, based on the information on the number of vehicles on road segments at different future times, we can more accurately predict the time required to travel to the target road segment at different times.

[0092] Furthermore, the server can determine, based on information from other routes, a first route that includes the target road segment and a second route that includes other road segments connected to the target road segment.

[0093] If the information for other routes includes the target road segment, it means that traveling along that route will lead to that target road segment. Therefore, vehicles traveling along that route will affect the road conditions of the target road segment when they reach it.

[0094] Based on this, a first route including the target road segment can be identified among other travel routes, and the first temporal distribution information of the target road segment in the future can be predicted based on the information of each first route. Similarly, a second route including other road segments connecting to the target road segment can be identified among other travel routes, and the second temporal distribution information of the other road segments in the future can be predicted based on the information of each second route.

[0095] In practical applications, the information for the first route includes the estimated time to reach the segments along the first route, and the information for the second route includes the estimated time to reach the segments along the second route. The information for a single route includes the time to reach segment L4 (t4), segment L5 (t5), and segment L6 (t6).

[0096] Specifically, the server can obtain the first estimated time to reach the target road segment from the information of the first route, and the second estimated time to reach other road segments from the information of the second route. For example, if there is a target road segment L4, the first estimated time t4 to reach L4 can be obtained from the information of the first route.

[0097] Specifically, based on the first estimated arrival times of each target road segment, the number of first vehicles arriving at the target road segment in each future time segment is determined. Based on the second estimated arrival times of each other road segment, the number of second vehicles arriving at other road segments in each future time segment is determined. For example, if multiple first routes include road segment L4, with arrival times of 10:55, 10:58, 11:10, 11:13, and 11:14, then the number of vehicles arriving at road segment L4 in the time segment from 10:55 to 11:00 is 2, and the number of vehicles arriving at road segment L4 in the time segment from 11:10 to 11:15 is 3.

[0098] Optionally, the predicted time for the target vehicle to arrive at the target road segment according to the target route can be determined, and the first vehicle that may arrive at the target road segment at the predicted time can be determined based on the first route. Then, the first time of arrival of these first vehicles at the target road segment can be obtained, thereby obtaining the first time distribution information.

[0099] It is also possible to determine the second vehicles that may arrive at other road segments at the predicted time based on the second route, and then obtain the second time when these second vehicles arrive at other road segments, thereby obtaining the second time distribution information.

[0100] It can be assumed that the first and second vehicles entering the target road segment after the target vehicle will not affect the travel time of the target vehicle through the target road segment. Therefore, this method can exclude the distance data of vehicles entering the target road segment and other road segments after the target vehicle, and can more accurately predict the travel time of the target road segment.

[0101] In this implementation, by statistically analyzing information from a large number of other travel routes, the number of vehicles passing through a road segment within a fixed time period can be determined. Since the number of vehicles on a road segment varies across different time periods, the time required to traverse that segment will also vary. Therefore, based on the travel routes already planned for users, the number of vehicles on a road segment in multiple future time periods can be predicted. Consequently, based on future road conditions, the time required to traverse that road segment within a given time period can be accurately estimated.

[0102] In this implementation, the first time distribution information is the number of the first vehicles arriving at the target road segment in each future time segment, and the second time distribution information is the number of the second vehicles arriving at other road segments in each future time segment. The specific number of time segments is N, which can be set according to requirements, for example, it could be 10 or 20. The length of each time segment can also be set according to requirements, for example, it could be 5 minutes or 3 minutes.

[0103] Step 304: Based on information from other travel routes, determine the first distance distribution information between the first vehicle that needs to pass through the target road segment and the target road segment, and the second distance distribution information between the second vehicle that needs to pass through other road segments and other road segments.

[0104] Furthermore, the time taken to traverse a road segment is also related to the number of other vehicles heading towards that segment when the user arrives. For example, if there are many other vehicles about to enter the target road segment when the user arrives, the time taken to traverse the target road segment may increase. Conversely, if the vehicles needing to enter the target road segment are all far away when the user arrives, the time taken to traverse the target road segment may decrease.

[0105] This method can determine the distribution of vehicles that need to reach a road segment from the distance dimension, thereby accurately estimating the time required to pass through the road segment using the distance distribution information.

[0106] Therefore, the solution provided in this disclosure also determines the first distance distribution information of the target road segment and the second distance distribution information of other road segments connected to the target road segment based on the target travel route.

[0107] The first distance distribution information is used to characterize the distance between the first vehicle that needs to enter the target road segment and the target road segment; the second distance distribution information is used to characterize the distance between the second vehicle that needs to enter other road segments connected to the target road segment and those other road segments.

[0108] In one optional implementation, the distance between the first vehicle needing to enter the target road segment and the target road segment at the current moment can be determined, or the distance between the first vehicle needing to enter the target road segment and the target road segment can be determined when vehicles traveling along the target route arrive at the target road segment. No specific restrictions are imposed. The distances between the second vehicles and other road segments are similar to the above and will not be elaborated further.

[0109] When determining distance distribution information, the server can determine a first route that includes the target road segment and a second route that includes other road segments connected to the target road segment, based on information from other routes.

[0110] If the actual application requires both time distribution information and distance distribution information, then it is sufficient to determine the first and second routes only once, without having to determine them repeatedly.

[0111] The server can obtain the first position of the first vehicle traveling along the first route, determine the first distance between the first vehicle and the target road segment based on the first position, and determine the first distance distribution information based on each first distance.

[0112] Optionally, the predicted time for the target vehicle to arrive at the target road segment according to the target route can be determined, and the first vehicle that may arrive at the target road segment at the predicted time can be determined based on the first route. Then, the current first position of these first vehicles can be obtained, and the first distance distribution information can be determined based on the first distance between each first vehicle and the target road segment when they may arrive at the target road segment at the predicted time. The second distance distribution information can also be obtained in a similar way.

[0113] It can be assumed that the first vehicle entering the target road segment after the target vehicle will not affect the travel time of the target vehicle through the target road segment. Therefore, this method can exclude the distance data of vehicles entering the target road segment after the target vehicle, and can more accurately predict the travel time of the target road segment.

[0114] The server can also obtain the second location of the second vehicle traveling along the second route, determine the second distance between the second vehicle and the target road segment based on the second location, and determine the second distance distribution information based on each second distance.

[0115] Furthermore, after determining the first location, the server can also determine the first distance between the first location and the target road segment, specifically the first distance between the first location and the starting point of the target road segment. Based on these first distances, the distribution information of the first distances corresponding to the target road segment can be determined.

[0116] For example, the first distance distribution information may include the number of vehicles 500 meters away from the target road segment, the number of vehicles 1000 meters away from the target road segment, etc.

[0117] Similarly, the second distance between the second location and other road segments can be determined, and the distribution information of the second distance can be obtained.

[0118] In one alternative implementation, if the determined position of a vehicle has exceeded the currently processed road segment, the distance between the vehicle and the road segment can be set to a negative number. For example, if the target road segment is L2, and the vehicle will first pass through road segment L2 and then road segment L3 when traveling along the route, if the estimated position of the vehicle is on road segment L3, the distance between the vehicle and the road segment can be set to a negative number.

[0119] The scheme disclosed herein determines a first distance distribution information and a second distance distribution information. These distance distribution information can characterize the distribution of vehicles that may reach the target road segment from the distance dimension. The vehicles that may enter the target road segment will affect the travel time of the target road segment. Therefore, by combining the distance distribution information, the future travel time of the target road segment can be predicted more accurately.

[0120] Step 305: Obtain the first speed of vehicles currently located on the target road segment, and the second speed of vehicles currently located on other road segments connected to the target road segment. The first road condition features include: first time distribution information, first distance distribution information, and first speed; the second road condition features include: second time distribution information, second distance distribution information, and second speed.

[0121] In addition, the time taken to traverse a road segment is also related to the speed of vehicles traveling on that target road segment. In the solution provided in this disclosure, the first speed of the vehicle currently located on the target road segment and the second speed of the vehicle currently located on other road segments connected to the target road segment can also be obtained.

[0122] For example, if we can obtain the real-time speeds of n1 vehicles located on the target road segment, we can determine the average of these real-time speeds to obtain the first speed. Similarly, if we can obtain the real-time speeds of n2 vehicles located on another road segment, we can determine the average of these real-time speeds to obtain the second speed of that other road segment.

[0123] In this disclosed solution, the first road condition characteristics include: first time distribution information, first distance distribution information, and first speed; the second road condition characteristics include: second time distribution information, second distance distribution information, and second speed. This information can characterize dynamic road condition information; therefore, using this information, the time required to traverse a road segment can be determined more accurately.

[0124] Step 306: Input the first road condition feature of the target road segment and the second road condition feature corresponding to the target road segment into the preset time prediction model to obtain the travel time of passing through the target road segment in each future time segment.

[0125] Specifically, in one optional implementation, a time prediction model can be set up, which can process the input traffic characteristics and output the travel time to the target road segment.

[0126] Furthermore, a time prediction model can be pre-set, and this model can be pre-trained. Specifically, it can be trained using training data to obtain the time prediction model. For example, route planning requests pre-cached before time t can be used to extract the features of road segments and the features of other road segments adjacent to that road segment. Historical logs can also be used to determine the actual travel time of that road segment in each time segment after time t. This actual travel time can then be used as supervision data, and the features of the road segment and the features of other road segments adjacent to that road segment can be used as input data to train the model. This allows the model to learn the ability to calculate the travel time of that road segment in each future time segment using the input traffic features.

[0127] Optionally, the extracted road segment features can include time-related features, such as the number of vehicles arriving at the road segment every 5 minutes. In this approach, the time prediction model can output the predicted travel duration of the target road segment within each time segment. For example, in the first time segment, the travel duration of the first target road segment is t11; in the second time segment, the travel duration of the first target road segment is t21; and in the first time segment, the travel duration of the second target road segment is t12. The length of this time segment can be, for example, 5 minutes.

[0128] Furthermore, after training the time prediction model, it can be deployed on a server. The target travel route includes multiple target road segments. The server can input the first traffic condition feature of a target road segment and the second traffic condition feature of other road segments connected to the target road segment into the deployed time prediction model to obtain the first estimated travel time of the target road segment.

[0129] In practical applications, the first road condition feature of the target road segment and the second road condition feature of other road segments can both include road condition information of the road segments at different time segments. For example, the first road condition feature includes the number of vehicles in the target road segment in the first time segment and the number of vehicles in the target road segment in the second time segment; the second road condition feature includes the number of vehicles in other road segments in the first time segment and the number of vehicles in other road segments in the second time segment.

[0130] In this implementation, when the training can output the travel time of road segments in different time segments, the road condition features in the training data also include the road condition information of road segments in different time segments, and the supervision data also includes the actual travel time of road segments in each time segment.

[0131] Since the road conditions of the target road segment will vary at different times, the corresponding travel time will also vary. Therefore, this method can more accurately predict the time required to travel through the target road segment after arriving at the target road segment when following the target route.

[0132] Step 307: Based on the departure time corresponding to the target route planning request, determine the target sub-duration in the travel sub-time of the first target road segment that needs to be passed.

[0133] The target route includes multiple target road segments, and the target sub-time of each target road segment can be determined one by one when traveling according to the target route.

[0134] For the first target road segment that needs to be traversed, the server can determine the target sub-duration from the travel sub-times of that target road segment. Specifically, the target sub-duration can be selected from various travel sub-durations based on the arrival time of the target road segment. For example, if the departure time is 15:00, the travel sub-time corresponding to the time segment from 15:00 to 15:05 can be determined as the target sub-duration from the travel sub-times of the target road segment.

[0135] Step 308: Determine the departure time of the next target road segment to be traversed based on the departure time and the target sub-duration, and determine the next target sub-duration from the travel sub-duration of the next target road segment.

[0136] Specifically, the server can add the target sub-duration to the departure time to obtain the time to reach the next target road segment, which is the departure time of the next target road segment.

[0137] Furthermore, the server can determine the next target sub-duration from among the multiple travel sub-durations of the target road segment, specifically the travel sub-duration corresponding to the time segment to which the departure time of the road segment belongs.

[0138] If the target route includes other target road segments, then step 308 can be executed to determine the departure time of the next target road segment and the next target sub-duration for passing through the next target road segment.

[0139] Step 309: Determine the total estimated travel time of the target route based on the target sub-time of each target road segment.

[0140] In practical applications, the server can overlay the target sub-time of each target road segment to obtain the total estimated travel time of the entire target route.

[0141] For example, if the target travel route includes target road segments L1, L2, and L3, then the target sub-times of L1, L2, and L3 can be determined separately, such as T1, T2, and T3 respectively. Then, T1, T2, and T3 can be superimposed to obtain the total estimated travel time of the target travel route.

[0142] In this implementation method, since the target sub-time of each target road segment is determined based on future road conditions and is consistent with the actual road conditions of the future target road segment, it is relatively accurate. Therefore, the total estimated travel time obtained by superimposing these target sub-times is also relatively accurate.

[0143] Figure 4 This is a schematic diagram illustrating a process for determining the travel time of each target road segment, as shown in an exemplary embodiment of this disclosure.

[0144] like Figure 4As shown, assume that the target route includes three target road segments: L1, L2, and L3, and that when traveling along the target route, one will pass through L1, L2, and L3 in sequence.

[0145] For the target road segment L1, the server can determine the travel time T1 of L1 at departure time t0 based on the information of other cached routes, and obtain the arrival time (t0+T1) of L2 based on departure time t0 and T1. Then, it can determine the travel time T2 of L2 at (t0+T1), determine the arrival time (t0+T1+T2) of L3 based on (t0+T1) and T2, and finally determine the arrival time T3 of L3 at (t0+T1+T2).

[0146] T1, T2, and T3 can be superimposed to obtain the total estimated travel time for the target route.

[0147] In one alternative implementation, multiple target routes can be planned for the target route planning request. In this case, the total estimated travel time can be determined for each target route based on the methods provided in this disclosure.

[0148] For example, if 10 target routes are planned, then 10 total estimated travel times can be determined for each of them.

[0149] In this scenario, the server can select N routes with the shortest total estimated travel time from multiple target routes. For example, it can select 3 routes with the shortest total estimated travel time, and then send the selected routes back to the user terminal. N is an integer greater than or equal to 1.

[0150] In this implementation, the server can push shorter travel routes to the user terminal to improve the user experience.

[0151] Figure 5 This is a schematic flowchart illustrating a model training method for estimating the travel time of a road segment, as shown in an exemplary embodiment of this disclosure.

[0152] like Figure 5 As shown, the model training method for estimating road segment travel time provided in this disclosure includes:

[0153] Step 501: Obtain the training dataset, which includes the first traffic condition features of the target road segment, the second traffic condition features of other road segments connected to the target road segment, and the first travel time corresponding to the target road segment; wherein, the first traffic condition features and the second traffic condition features are determined based on historical planned routes.

[0154] Specifically, the solutions provided in this disclosure can be applied to electronic devices with computing capabilities, such as computers.

[0155] Furthermore, historical planned routes can be obtained, such as those from the previous day, and the model can be trained using these historical planned routes. Historical planned routes are routes actually planned by the server based on navigation requests.

[0156] In practical applications, the first road condition features of the target road segment and the second road condition features of other road segments connected to the target road segment can be extracted based on the historical planned route. For example, the first road condition features of the target road segment after the first moment can be predicted based on the historical planned route before the first moment, as well as the second road condition features of other road segments connected to the target road segment.

[0157] The target road segment can be any road segment. Based on the historical planned routes including the target road segment generated before the first moment, the road conditions of the target road segment after the first moment can be predicted. For example, if the historical planned routes are data from one day ago, the first road condition feature of the target road segment at 10:00 can be extracted using the historical planned routes before 10:00, as well as the second road condition features of other road segments connected to the target road segment.

[0158] Specifically, historical planned routes can also have a validity period. For example, if a user sends a navigation request through their terminal device, and the server provides the planned route to the user's terminal, and the user drives along that route, then the validity period of that route is from the time it was generated to the time the user exits the navigation. On the other hand, if the user exits the navigation software after the route is displayed on their terminal, then the validity period of that route is shorter.

[0159] Specifically, the server can use historical planned routes generated before the first moment and with an expiration date later than the first moment to extract the first road condition features of the target road segment at the first moment, as well as the second road condition features of other road segments connected to the target road segment.

[0160] The first road condition feature is used to characterize the road condition information of the target road segment extracted at the first moment, such as the number of vehicles that may arrive on the target road segment in each time period after the first moment, or the average speed of vehicles on the target road segment.

[0161] Furthermore, other road segments may connect to the target road segment, and the road conditions of these other road segments typically affect the road conditions of the target road segment. Therefore, the method provided in this disclosure can also determine second road condition features of other road segments connected to the target road segment based on information from other travel routes. These second road condition features are used to characterize the road condition information of the other road segments extracted at the first time point.

[0162] In practical applications, the server can also obtain historical log data and use it to obtain the first passage time corresponding to the target road segment.

[0163] For example, the time it takes for a vehicle to pass through a target road segment can be determined based on historical log data. Historical log data may include the speed of vehicles located on the target road segment at various times. The time it takes for each vehicle to pass through the target road segment can be determined based on the speed information and the travel time of the target road segment. The average of these times is then determined as the first travel time of the target road segment.

[0164] In one optional implementation, since the travel time of the target road segment will also be different in different time periods, the travel time of the target road segment in different time segments can also be determined based on historical log data of different time segments. For example, the travel time of target road segment 1 in the time segment of 20:00-20:05, and the travel time of target road segment 1 in the time segment of 20:05-20:10.

[0165] Step 502: Input the first road condition feature and the second road condition feature into the preset model to obtain the predicted travel time corresponding to the target road segment.

[0166] The model can be pre-set. The server can input the first road condition feature and the second road condition feature of the target road segment into the pre-set model. The pre-set model can output the predicted travel time required to pass through the target road segment.

[0167] Specifically, the road conditions of the target road segment will be different at different times. Therefore, the first road condition feature of the target road segment extracted at the same time, and the second road condition feature of other road segments connected to the target road segment at that time, can be input into the preset model. The preset model can output the predicted travel time of the target road segment based on the first and second road condition features at that time.

[0168] Step 503: Adjust the parameters in the model according to the first travel time and the predicted travel time to obtain a model for estimating the travel time of road segments.

[0169] Furthermore, the server can construct a loss function based on the first travel time of the target road segment and the predicted travel time of the target road segment, and then use the loss function for gradient backpropagation to adjust the parameters in the model.

[0170] In practical applications, multiple iterations can be performed to adjust the parameters in the model multiple times, resulting in a model that can accurately output the travel time of road segments.

[0171] In one alternative implementation, the predicted travel time output by the model may include travel sub-times corresponding to multiple time periods. For example, by using the first road condition features of the target road segment at the first moment and the second road condition features of other road segments as inputs into the model, the travel sub-times predicted for each time segment after the first moment can be obtained.

[0172] In this implementation, the first travel duration of the target road segment, which serves as the monitoring data, may also include the travel sub-duration corresponding to each time segment.

[0173] The trained model can be deployed on a server, enabling the server to perform tasks such as... Figure 1-4 Any of the embodiments shown.

[0174] The model training method for estimating travel time on road segments provided in this disclosure can extract features that characterize future road conditions based on information from historical planned routes, and then obtain the actual travel time required to pass through the target road segment using historical logs. By training the model with this data, the model can learn to determine the travel time required to pass through the target road segment based on the road condition features, and thus obtain a more accurate travel time required to pass through the target road segment.

[0175] Figure 6 This is a schematic flowchart illustrating a model training method for estimating the travel time of a road segment, as shown in another exemplary embodiment of this disclosure.

[0176] like Figure 6 As shown, the model training method for estimating road segment travel time provided in this disclosure includes:

[0177] Step 601: Obtain the historical planned routes generated based on the route planning request before the preset time.

[0178] The server can obtain a large number of route planning requests and historical logs. Specifically, it can extract road condition features based on route planning requests before a preset time, determine the travel time of each target road segment after the preset time based on historical logs, and use this data to train the model, enabling the model to learn the ability to output the travel time of the target road segment based on these road condition features.

[0179] Step 602: Based on the information of historical planned routes, determine the first time distribution information of the number of vehicles arriving at the target road segment after a preset time, and the second time distribution information of the number of vehicles arriving at other road segments after a preset time.

[0180] The information on the historical planned route includes the road segments along the route, and may also include the time taken to reach each road segment when traveling along the historical planned route. For example, the time to reach each road segment can be determined based on the method of this disclosure, or it can be determined based on existing technology.

[0181] Specifically, by analyzing the arrival times of each road segment in the historical route planning information, we can obtain the first time distribution information of the target road segment, as well as the second time distribution information of other road segments connected to the target road segment.

[0182] Temporal distribution information is used to characterize the number of vehicles arriving at a road segment at different times, that is, to determine the distribution of vehicles arriving at a road segment according to the time dimension.

[0183] Furthermore, the computer can determine a first route that includes the target road segment and a second route that includes other road segments connected to the target road segment based on information from historical planned routes.

[0184] If the historical route planning information includes a target road segment, it means that when traveling along the route, the vehicle will reach that target road segment. Therefore, when a vehicle traveling along the route reaches the target road segment, it will affect the road conditions of the target road segment.

[0185] Based on this, a first route including the target road segment can be identified from the historical planned routes, and the first time distribution information of the target road segment after a preset time can be determined based on the information of each first route. Similarly, a second route including other road segments connected to the target road segment can be identified from the historical planned routes, and the second time distribution information of the other road segments after a preset time can be determined based on the information of each second route.

[0186] In practical applications, the information for the first route includes the estimated time to reach the segments along the first route, and the information for the second route includes the estimated time to reach the segments along the second route. The information for a single route includes the time to reach segment L4 (t4), segment L5 (t5), and segment L6 (t6).

[0187] Specifically, the computer obtains a first estimated time to reach the target road segment from the information of the first route, and a second estimated time to reach other road segments from the information of the second route. Based on the first estimated times to reach the target road segment, the computer determines the number of first vehicles arriving at the target road segment within each time segment after a preset time, and based on the second estimated times to reach other road segments, the computer determines the number of second vehicles arriving at other road segments within each time segment after the preset time.

[0188] Optionally, preset times can be set. Based on the first route, the first vehicles that may arrive at the target road segment at the preset times can be determined, and then the arrival times of these first vehicles at the target road segment can be obtained, thus obtaining the first time distribution information. Second time distribution information can also be obtained in a similar way.

[0189] It can be assumed that the first vehicle entering the target road segment after the preset time will not affect the travel time of vehicles that have traveled through the target road segment before the preset time. Therefore, this method can exclude the distance data of vehicles that entered the target road segment before the preset time when training the model based on the actual travel time of the target road segment after the preset time, thus training a more accurate model. For example, the road condition features at the preset time can be input into the model, so that the model outputs the predicted travel time after the preset time. This predicted travel time includes the predicted travel sub-time corresponding to the preset time, and then the model parameters are adjusted according to the actual travel time corresponding to the preset time.

[0190] The first time distribution information is the number of the first vehicles arriving at the target road segment within each future time segment, and the second time distribution information is the number of the second vehicles arriving at other road segments within each future time segment. The specific number of time segments is N, which can be set according to requirements, for example, it could be 10 or 20. The length of each time segment can also be set according to requirements, for example, it could be 5 minutes or 3 minutes.

[0191] Step 603: Based on the information of the historical planned routes, determine the first distance distribution information between the first vehicle that needs to pass through the target road segment and the target road segment, and the second distance distribution information between the second vehicle that needs to pass through other road segments and other road segments.

[0192] This method can determine the distribution of vehicles that need to reach a road segment from the distance dimension, thereby accurately estimating the time required to pass through the road segment using the distance distribution information.

[0193] The first distance distribution information is used to represent the distance between the first vehicle needing to enter the target road segment and the target road segment; the second distance distribution information is used to represent the distance between the second vehicle needing to enter another road segment connected to the target road segment and that other road segment. For example, there are 10 vehicles 500 meters away from the target road segment and 50 vehicles 1000 meters away from the target road segment.

[0194] Specifically, based on information from historical planned routes, a first route including the target road segment and a second route including other road segments connecting to the target road segment can be determined from the historical planned routes.

[0195] If the actual application requires both time distribution information and distance distribution information, then it is sufficient to determine the first and second routes only once, without having to determine them repeatedly.

[0196] Obtain the first position of the first vehicle traveling along the first route at a preset time, determine the first distance between the first vehicle and the target road segment based on the first position, and determine the first distance distribution information based on each first distance;

[0197] Based on the second route, determine the second position of the second vehicle that needs to pass through other road sections at a preset time, and determine the second distance between the second vehicle and the target road section based on the second position, and determine the second distance distribution information based on each second distance.

[0198] The server can obtain the first position of the first vehicle traveling along the first route, determine the first distance between the first vehicle and the target road segment based on the first position, and determine the first distance distribution information based on each first distance.

[0199] Optionally, a preset time can be set, and the first vehicles that may arrive at the target road segment before the preset time can be determined based on the first route. Then, the first positions of these first vehicles can be obtained, and the first distance distribution information can be determined based on the first distance between each first vehicle and the target road segment when it may arrive at the target road segment before the preset time. Second distance distribution information can also be obtained in a similar way.

[0200] It can be assumed that the first vehicle entering the target road segment after the preset time will not affect the travel time of the vehicle that passed through the target road segment before the preset time. Therefore, this method can train the model more accurately.

[0201] In one optional implementation, the distance between the first vehicle needing to enter the target road segment and the target road segment at the current moment can be determined, or the distance between the first vehicle needing to enter the target road segment and the target road segment can be determined when vehicles traveling along the target route arrive at the target road segment. No specific restrictions are imposed. The distances between the second vehicles and other road segments are similar to the above and will not be elaborated further.

[0202] Step 604: Based on the historical logs, determine the first speed of the vehicle located on the target road segment and the second speed of the second vehicle located on other road segments connected to the target road segment when the historical planned route was generated.

[0203] When a user terminal is running navigation software, it can report speed information. Therefore, the computer can access historical logs to obtain the first speed of vehicles on the target road segment and the second speed of vehicles on other road segments.

[0204] Specifically, the vehicles located on the target road segment and other road segments are different at different times. Therefore, the first and second speeds obtained can also contain time information. For example, it can include the first speed of the first vehicle on the target road at 10:00 and the speed of the first vehicle on the target road at 11:00.

[0205] In this disclosed scheme, the first road condition feature includes: first time distribution information, first distance distribution information, and first speed; the second road condition feature includes: second time distribution information, second distance distribution information, and second speed.

[0206] Step 605: Determine the first passage duration for the target road segment within the time segment based on the historical logs; the first passage duration includes the first passage duration for multiple time segments.

[0207] Specifically, historical logs are used to determine the actual travel time of the road segment within each time slot, and this actual travel time is then used as supervisory data. Based on the historical logs, the time it takes for a vehicle to pass through the target road segment at different time slots can be determined, thus obtaining the first travel sub-durations of the target road segment. For example, if a vehicle arrives at the target road segment at 10:00 and its travel time is 10 minutes, and arrives at the target road segment at 10:05 and its travel time is 15 minutes, then the first travel sub-duration corresponding to the time slot of 10:00-10:05 can be determined as the average of all travel times corresponding to that time slot, which is 12.5 minutes.

[0208] Step 606: Obtain the training dataset, which includes the first traffic condition features of the target road segment, the second traffic condition features of other road segments connected to the target road segment, and the first travel time corresponding to the target road segment; wherein, the first traffic condition features and the second traffic condition features are determined based on historical planned routes.

[0209] Step 606 is implemented in a similar way to step 51, and will not be described again.

[0210] Step 607: Input the first time distribution information, first distance distribution information, and first speed of the target road segment, as well as the second time distribution information, second distance distribution information, and second speed of other road segments connected to the target road segment, into the preset model to obtain the predicted travel time through the target road segment in each time segment after the preset time.

[0211] Optionally, the extracted road segment features may include time-related features, first time distribution information, and second time distribution information. In this approach, the pre-defined model can output the predicted travel duration of the target road segment within each time segment. For example, in the first time segment, the predicted travel duration t11 for the first target road segment; in the second time segment, the predicted travel duration t21 for the first target road segment; and in the first time segment, the predicted travel duration t12 for the second target road segment. The length of this time segment can be, for example, 5 minutes.

[0212] In practical applications, the road condition features input to the model are determined based on the road condition features after a preset time, according to the historical planned routes. Therefore, the solution provided in this disclosure can train the model and determine the predicted travel time of the target road segment in different time segments after the preset time, based on the road condition features after the preset time.

[0213] Step 608: Adjust the parameters in the model according to the first travel time and each predicted travel time to obtain a model for estimating the travel time of road segments.

[0214] The first travel time can include the actual travel time through the target road segment in different time segments, and the predicted travel time also includes the predicted travel time through the target road segment output by the model in different time segments.

[0215] Specifically, the computer can compare the actual duration with the predicted duration within the same time segment, and then adjust the model parameters based on the comparison results. Through multiple iterations, the predicted duration output by the model can be made close to the actual duration, thus obtaining a model for estimating the travel time of road segments.

[0216] Figure 7 This is a schematic diagram of the structure of a travel time estimation device shown in an exemplary embodiment of the present disclosure.

[0217] like Figure 7 As shown, the travel time estimation device 700 provided in this disclosure includes:

[0218] The route acquisition unit 710 is used to acquire the target route corresponding to the target route planning request, wherein the target route includes multiple target road segments.

[0219] The route acquisition unit 710 is also used to acquire information on other routes corresponding to other route planning requests;

[0220] The feature extraction unit 720 is used to determine the first road condition features of the target road segment and the second road condition features of other road segments connected to the target road segment based on the information of the other travel routes.

[0221] The road segment travel time determination unit 730 is used to determine the first estimated travel time of the target road segment based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment.

[0222] The route travel time determination unit 740 is used to determine the total estimated travel time of the target travel route based on the first estimated travel time of each target road segment.

[0223] The travel time estimation device provided in this disclosure can predict the future road conditions of each target road segment based on existing information of other travel routes, and then predict the first estimated travel time required to travel the target road segment based on the road conditions of each target road segment. This method of estimating the time to travel through the target road segment based on predicted road condition information can more accurately estimate the time required to travel through the target road segment in the future, and thus obtain a more accurate total estimated travel time required to travel through the target route.

[0224] Figure 8 This is a schematic diagram of the structure of a travel time estimation device shown as another exemplary embodiment of the present disclosure.

[0225] like Figure 8 As shown, in the travel time estimation device 800 provided in this disclosure, the route acquisition unit 810 and Figure 7 The route acquisition unit 710 shown is similar to the feature extraction unit 820. Figure 7 The feature extraction unit 720 shown is similar to the road segment travel time determination unit 830. Figure 7 The road segment travel time determination unit 730 shown is similar to the route travel time determination unit 840. Figure 7 The route travel time determination unit 740 shown is similar.

[0226] The feature extraction unit 820 includes:

[0227] The time distribution information extraction module 821 is used to determine, based on the information of the other travel routes, a first time distribution information of the number of vehicles arriving at the target road segment in the future and a second time distribution information of the number of vehicles arriving at the other road segments in the future;

[0228] The distance distribution information extraction module 822 is used to determine, based on the information of the other travel routes, the first distance distribution information between the first vehicle that needs to pass through the target road segment and the target road segment, and the second distance distribution information between the second vehicle that needs to pass through the other road segments and the other road segments;

[0229] The speed information extraction module 823 is used to obtain the first speed of the vehicle currently located on the target road segment, and the second speed of the vehicle currently located on other road segments connected to the target road segment;

[0230] The first road condition feature includes: first time distribution information, first distance distribution information, and first speed; the second road condition feature includes: second time distribution information, second distance distribution information, and second speed.

[0231] Specifically, the time distribution information extraction module 821 is used for:

[0232] Based on the information of the other routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined among the other routes;

[0233] In the information of the first route, a first estimated time to reach the target road segment is obtained, and in the information of the second route, a second estimated time to reach the other road segments is obtained; wherein, the information of the first route includes the estimated time to reach the road segments passed through in the first route, and the information of the second route includes the estimated time to reach the road segments passed through in the second route.

[0234] Based on the first estimated time of arrival at the target road segment, the number of first vehicles arriving at the target road segment in each future time segment is determined; based on the second estimated time of arrival at the other road segments, the number of second vehicles arriving at the other road segments in each future time segment is determined; wherein, the first time distribution information is the number of first vehicles arriving at the target road segment in each future time segment, and the second time distribution information is the number of second vehicles arriving at the other road segments in each future time segment.

[0235] The distance distribution information extraction module 822 is specifically used for:

[0236] Based on the information of the other routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined among the other routes;

[0237] Obtain the first position of the first vehicle traveling along the first route, determine the first distance between the first vehicle and the target road segment based on the first position, and then determine the first distance distribution information based on each of the first distances;

[0238] Obtain the second position of the second vehicle traveling along the second route, determine the second distance between the second vehicle and the target road segment based on the second position, and determine the second distance distribution information based on each of the second distances.

[0239] The road segment travel time determination unit 830 is specifically used for:

[0240] The first road condition feature of the target road segment and the second road condition feature corresponding to the target road segment are input into a preset time prediction model to obtain the travel time of passing through the target road segment in each future time segment.

[0241] Wherein, the first estimated travel time includes the travel sub-time for passing through the target road segment within each time segment;

[0242] The route travel time determination unit 840 includes:

[0243] The target sub-duration determination module 841 is used to determine the target sub-duration based on the departure time corresponding to the target route planning request, from the travel sub-duration of the first target road segment that needs to be passed.

[0244] The target sub-duration determination module 841 is further configured to determine the departure time of the next target road segment to be passed based on the departure time and the target sub-duration, and to determine the next target sub-duration in the passage sub-duration of the next target road segment;

[0245] The total duration determination module 842 is used to determine the total estimated travel time of the target route based on the target sub-duration of each target road segment.

[0246] The target route planning request includes multiple target routes;

[0247] The device further includes a screening unit 850, used for:

[0248] Based on the total estimated travel time of each target route, select the N routes with the shortest total estimated travel time from the target routes and return the N routes; N is an integer greater than or equal to 1.

[0249] Figure 9 This is a schematic diagram of the structure of a model training device for estimating the travel time of a road segment, as shown in an exemplary embodiment of the present disclosure.

[0250] like Figure 9 As shown, the model training device 900 for estimating the travel time of road segments provided in this disclosure includes:

[0251] The training data acquisition unit 910 is used to acquire a training dataset, which includes a first traffic condition feature of the target road segment, a second traffic condition feature of other road segments connected to the target road segment, and a first travel time corresponding to the target road segment; wherein the first traffic condition feature and the second traffic condition feature are determined based on historical planned routes.

[0252] The prediction unit 920 is used to input the first road condition feature and the second road condition feature into a preset model to obtain the predicted travel time corresponding to the target road segment;

[0253] The adjustment unit 930 is used to adjust the parameters in the model according to the first travel time and the predicted travel time to obtain a model for estimating the travel time of the road segment.

[0254] The model training device for estimating the travel time of a road segment provided in this disclosure can extract features that can characterize future road conditions based on information from historical planned routes, and then obtain the actual time required to pass through the target road segment using historical logs. By training the model with this data, the model can learn to determine the travel time required to pass through the target road segment based on the road condition features, and thus obtain a more accurate travel time required to pass through the target road segment.

[0255] Figure 10 This is a schematic diagram of the structure of a model training device for estimating the travel time of a road segment, as shown in another exemplary embodiment of this disclosure.

[0256] like Figure 10 As shown, in the model training device 1000 for estimating road segment travel time provided in this disclosure, the training data acquisition unit 1010 and... Figure 9 The training data acquisition unit 910 is similar to the prediction unit 1020. Figure 9 The prediction unit 920 is similar to the adjustment unit 1030. Figure 9 It is similar to the adjustment unit 930 in the middle.

[0257] The device further includes a data preprocessing unit 1040, used for:

[0258] Obtain historical planned routes generated based on route planning requests prior to the preset time;

[0259] Based on the information of the historical planned routes, determine the first time distribution information of the number of vehicles arriving at the target road segment after the preset time, and the second time distribution information of the number of vehicles arriving at the other road segments after the preset time;

[0260] Based on the information of the historical planned routes, determine the first distance distribution information between the first vehicle that needs to pass through the target road segment and the target road segment, and the second distance distribution information between the second vehicle that needs to pass through the other road segments and the other road segments;

[0261] Based on historical logs, determine the first speed of vehicles located on the target road segment, and the second speed of vehicles located on other road segments connected to the target road segment when the historical planned route was generated;

[0262] The first road condition feature includes: first time distribution information, first distance distribution information, and first speed; the second road condition feature includes: second time distribution information, second distance distribution information, and second speed.

[0263] The data preprocessing unit 1040 includes a time distribution information extraction module 1041, used for:

[0264] Based on the information of the historical planned routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined in the historical planned routes;

[0265] In the information of the first route, a first estimated time to reach the target road segment is obtained, and in the information of the second route, a second estimated time to reach the other road segments is obtained; wherein, the information of the first route includes the estimated time to reach the road segments passed through in the first route, and the information of the second route includes the estimated time to reach the road segments passed through in the second route.

[0266] Based on the first estimated time of arrival at the target road segment, determine the number of first vehicles arriving at the target road segment in each time segment after the preset time; based on the second estimated time of arrival at the other road segments, determine the number of second vehicles arriving at the other road segments in each time segment after the preset time.

[0267] The data preprocessing unit 1040 includes a distance information extraction module 1042, used for:

[0268] Based on the information of the historical planned routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined in the historical planned routes;

[0269] Obtain the first position of the first vehicle traveling along the first route at the preset time, determine the first distance between the first vehicle and the target road segment based on the first position, and determine the first distance distribution information based on each of the first distances;

[0270] Based on the second route, determine the second position of the second vehicle that needs to pass through the other road segments at the preset time, and determine the second distance between the second vehicle and the target road segment based on the second position, and determine the second distance distribution information based on each of the second distances.

[0271] Specifically, the prediction unit 1020 is used for:

[0272] The first time distribution information, the first distance distribution information, and the first speed of the target road segment, as well as the second time distribution information, the second distance distribution information, and the second speed of the other road segments connected to the target road segment, are input into a preset model to obtain the predicted travel time through the target road segment in each time segment after the preset time.

[0273] The predicted travel time includes the travel sub-time for passing through the target road segment within each time segment;

[0274] The data preprocessing unit 1040 further includes a supervisory data determination module 1043, used for:

[0275] The first passage duration for traveling the target road segment within the time segment is determined based on historical logs; the first passage duration includes the first passage duration for multiple time segments.

[0276] This disclosure provides a method for predicting travel time, a model training method, an electronic device, and a program product, which are applied to autonomous driving technology, intelligent transportation technology, vehicle networking technology, and intelligent cockpit technology in artificial intelligence technology, to accurately predict the travel time required to travel through a planned route.

[0277] It should be noted that the travel routes in this embodiment are not data specific to any particular user and do not reflect the personal information of any particular user.

[0278] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0279] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0280] According to embodiments of this disclosure, this disclosure also provides a computer program product comprising: a computer program stored in a readable storage medium, at least one processor of an electronic device being able to read the computer program from the readable storage medium, and the at least one processor executing the computer program causing the electronic device to perform the scheme provided in any of the above embodiments.

[0281] Figure 11 A schematic block diagram of an example electronic device 1100 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0282] like Figure 11As shown, device 1100 includes a computing unit 1101, which can perform various appropriate actions and processes according to a computer program stored in read-only memory (ROM) 1102 or a computer program loaded from storage unit 1108 into random access memory (RAM) 1103. The RAM 1103 may also store various programs and data required for the operation of device 1100. The computing unit 1101, ROM 1102, and RAM 1103 are interconnected via bus 1104. Input / output (I / O) interface 1105 is also connected to bus 1104.

[0283] Multiple components in device 1100 are connected to I / O interface 1105, including: input unit 1106, such as keyboard, mouse, etc.; output unit 1107, such as various types of monitors, speakers, etc.; storage unit 1108, such as disk, optical disk, etc.; and communication unit 1109, such as network card, modem, wireless transceiver, etc. Communication unit 1109 allows device 1100 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0284] The computing unit 1101 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1101 performs the various methods and processes described above, such as methods for estimating travel time or methods for training models for estimating travel time on road segments. For example, in some embodiments, the methods for estimating travel time or methods for training models for estimating travel time on road segments can be implemented as computer software programs tangibly contained in a machine-readable medium, such as storage unit 1108. In some embodiments, part or all of the computer program can be loaded and / or installed on device 1100 via ROM 1102 and / or communication unit 1109. When the computer program is loaded into RAM 1103 and executed by computing unit 1101, one or more steps of the travel time estimation method or the model training method for estimating road segment travel time described above can be performed. Alternatively, in other embodiments, computing unit 1101 can be configured by any other suitable means (e.g., by means of firmware) to perform the travel time estimation method or the model training method for estimating road segment travel time.

[0285] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0286] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0287] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0288] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0289] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0290] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service ecosystem, addressing the shortcomings of traditional physical hosts and VPS (Virtual Private Server, or simply "VPS") services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0291] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0292] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for estimating travel time, comprising: Obtain the target route corresponding to the target route planning request, wherein the target route includes multiple target road segments along the route; Information on other routes corresponding to other route planning requests is obtained, and a first road condition feature of the target road segment and a second road condition feature of other road segments connected to the target road segment are determined based on the information of the other routes. The first road condition feature includes a first distance distribution information and a first speed, and the second road condition feature includes a second distance distribution information and a second speed. The first distance distribution information is used to characterize the distance between a first vehicle needing to enter the target road segment and the target road segment. The second distance distribution information is used to characterize the distance between a second vehicle needing to enter other road segments connected to the target road segment and the other road segments. The first speed is the speed of a vehicle currently located on the target road segment, and the second speed is the speed of a vehicle currently located on other road segments connected to the target road segment. Based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment, the first estimated travel time of the target road segment is determined; The first estimated travel time includes the travel sub-time for passing through the target road segment in each future time segment; Based on the first estimated travel time of each target road segment, the total estimated travel time of the target travel route is determined; The acquisition of the first distance distribution information and the second distance distribution information includes: Based on the information of the other routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined among the other routes; Obtain the first position of the first vehicle traveling along the first route, determine the first distance between the first vehicle and the target road segment based on the first position, and then determine the first distance distribution information based on each of the first distances; Obtain the second position of the second vehicle traveling along the second route, determine the second distance between the second vehicle and the target road segment based on the second position, and determine the second distance distribution information based on each of the second distances.

2. The method according to claim 1, wherein, The first road condition feature further includes: first time distribution information; the second road condition feature further includes: second time distribution information, wherein determining the first road condition feature of the target road segment and the second road condition feature of other road segments connected to the target road segment based on the information of the other travel routes includes: Based on the information of the other travel routes, determine the first time distribution information of the number of vehicles arriving at the target road segment in the future, and the second time distribution information of the number of vehicles arriving at the other road segments in the future; The method further includes: Obtain the first speed and the second speed.

3. The method according to claim 2, wherein, The step of determining, based on information about other travel routes, the first time distribution information of the number of vehicles arriving at the target road segment in the future, and the second time distribution information of the number of vehicles arriving at other road segments in the future, includes: Based on the information of the other routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined among the other routes; In the information of the first route, a first estimated time to reach the target road segment is obtained, and in the information of the second route, a second estimated time to reach the other road segments is obtained; wherein, the information of the first route includes the estimated time to reach the road segments passed through in the first route, and the information of the second route includes the estimated time to reach the road segments passed through in the second route. Based on the first estimated time of arrival at the target road segment, the number of first vehicles arriving at the target road segment in each future time segment is determined; based on the second estimated time of arrival at the other road segments, the number of second vehicles arriving at the other road segments in each future time segment is determined; wherein, the first time distribution information is the number of first vehicles arriving at the target road segment in each future time segment, and the second time distribution information is the number of second vehicles arriving at the other road segments in each future time segment.

4. The method according to any one of claims 1-3, wherein, The step of determining the first estimated travel time of the target road segment based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment includes: The first road condition feature of the target road segment and the second road condition feature corresponding to the target road segment are input into a preset time prediction model to obtain the travel time of passing through the target road segment in each future time segment.

5. The method according to any one of claims 1-3, The step of determining the total estimated travel time of the route based on the first estimated travel time of each target road segment includes: Based on the departure time corresponding to the target route planning request, the target sub-time is determined from the travel sub-time of the first target road segment that needs to be traversed; The departure time of the next target road segment to be traversed is determined based on the departure time and the target sub-duration, and the next target sub-duration is determined from the travel sub-duration of the next target road segment; The total estimated travel time of the target route is determined based on the target sub-time of each target road segment.

6. The method according to any one of claims 1-3, wherein, The target route planning request has multiple target routes; The method further includes: Based on the total estimated travel time of each target route, select the N routes with the shortest total estimated travel time from the target routes and return the N routes; N is an integer greater than or equal to 1.

7. A model training method for predicting travel time on road segments, comprising: A training dataset is obtained, comprising a first traffic condition feature of a target road segment, second traffic condition features of other road segments connected to the target road segment, and a first travel time corresponding to the target road segment; wherein the first and second traffic condition features are determined based on historical planned routes; the first traffic condition feature includes a first distance distribution information and a first speed, and the second traffic condition feature includes a second distance distribution information and a second speed; the first distance distribution information is used to characterize the distance between a first vehicle needing to enter the target road segment and the target road segment; the second distance distribution information is used to characterize the distance between a second vehicle needing to enter other road segments connected to the target road segment and the other road segments; the first speed is the speed of a vehicle located on the target road segment, and the second speed is the speed of a vehicle located on other road segments connected to the target road segment when the historical planned route was generated; The first road condition feature and the second road condition feature are input into a preset model to obtain the predicted travel time corresponding to the target road segment; the predicted travel time includes the travel sub-time of passing through the target road segment in each time segment after the preset time. The parameters in the model are adjusted according to the first travel duration and the predicted travel duration to obtain a model for estimating the travel duration of road segments. The acquisition of the first distance distribution information and the second distance distribution information includes: Based on the information of the historical planned routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined in the historical planned routes; Obtain the first position of the first vehicle traveling along the first route at the preset time, determine the first distance between the first vehicle and the target road segment based on the first position, and determine the first distance distribution information based on each of the first distances; Based on the second route, determine the second position of the second vehicle that needs to pass through the other road segments at the preset time, and determine the second distance between the second vehicle and the target road segment based on the second position, and determine the second distance distribution information based on each of the second distances.

8. The method according to claim 7, further comprising: Obtain historical planned routes generated based on route planning requests prior to the preset time; Based on the information of the historical planned routes, determine the first time distribution information of the number of vehicles arriving at the target road segment after the preset time, and the second time distribution information of the number of vehicles arriving at the other road segments after the preset time; Based on historical logs, determine the first speed and the second speed; The first road condition feature further includes: first time distribution information; the second road condition feature further includes: second time distribution information.

9. The method according to claim 8, wherein, The step of determining, based on the information of the historical planned routes, the first time distribution information of the number of vehicles arriving at the target road segment after the preset time, and the second time distribution information of the number of vehicles arriving at other road segments after the preset time, includes: Based on the information of the historical planned routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined in the historical planned routes; In the information of the first route, a first estimated time to reach the target road segment is obtained, and in the information of the second route, a second estimated time to reach the other road segments is obtained; wherein, the information of the first route includes the estimated time to reach the road segments passed through in the first route, and the information of the second route includes the estimated time to reach the road segments passed through in the second route. Based on the first estimated time of arrival at the target road segment, determine the number of first vehicles arriving at the target road segment in each time segment after the preset time; based on the second estimated time of arrival at the other road segments, determine the number of second vehicles arriving at the other road segments in each time segment after the preset time.

10. The method according to claim 8, wherein, The step of classifying the first road condition feature and the second road condition feature into a preset model to obtain the predicted travel time corresponding to the target road segment includes: The first time distribution information, the first distance distribution information, and the first speed of the target road segment, as well as the second time distribution information, the second distance distribution information, and the second speed of the other road segments connected to the target road segment, are input into a preset model to obtain the predicted travel time through the target road segment in each time segment after the preset time.

11. The method according to any one of claims 7-10, The method further includes: The first passage duration for traveling the target road segment within the time segment is determined based on historical logs; the first passage duration includes the first passage duration for multiple time segments.

12. A device for estimating travel time, comprising: The route acquisition unit is used to acquire the target route corresponding to the target route planning request, wherein the target route includes multiple target road segments. The route acquisition unit is also used to acquire information on other routes corresponding to other route planning requests; A feature extraction unit is configured to determine, based on information from other travel routes, a first road condition feature of the target road segment and a second road condition feature of other road segments connected to the target road segment; the first road condition feature includes first distance distribution information and a first speed, and the second road condition feature includes second distance distribution information and a second speed; the first distance distribution information is used to characterize the distance between a first vehicle needing to enter the target road segment and the target road segment; the second distance distribution information is used to characterize the distance between a second vehicle needing to enter other road segments connected to the target road segment and the other road segments; the first speed is the speed of a vehicle currently located on the target road segment, and the second speed is the speed of a vehicle currently located on other road segments connected to the target road segment; The road segment travel time determination unit is used to determine the first estimated travel time of the target road segment based on the first road condition characteristics of the target road segment and the second road condition characteristics corresponding to the target road segment; The first estimated travel time includes the travel sub-time for passing through the target road segment in each future time segment; The route travel time determination unit is used to determine the total estimated travel time of the target travel route based on the first estimated travel time of each target road segment; The feature extraction unit includes a distance distribution information extraction module; The distance distribution information extraction module is used to determine, based on the information of the other routes, a first route including the target road segment and a second route including other road segments connected to the target road segment; Obtain the first position of the first vehicle traveling along the first route, determine the first distance between the first vehicle and the target road segment based on the first position, and then determine the first distance distribution information based on each of the first distances; Obtain the second position of the second vehicle traveling along the second route, determine the second distance between the second vehicle and the target road segment based on the second position, and determine the second distance distribution information based on each of the second distances.

13. The apparatus according to claim 12, wherein, The feature extraction unit includes: The time distribution information extraction module is used to determine, based on the information of the other travel routes, the first time distribution information of the number of vehicles arriving at the target road segment in the future and the second time distribution information of the number of vehicles arriving at the other road segments in the future; A speed information extraction module is used to obtain the first speed and the second speed; The first road condition feature further includes: first time distribution information; the second road condition feature further includes: second time distribution information.

14. The apparatus according to claim 13, wherein, The time distribution information extraction module is specifically used for: Based on the information of the other routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined among the other routes; In the information of the first route, a first estimated time to reach the target road segment is obtained, and in the information of the second route, a second estimated time to reach the other road segments is obtained; wherein, the information of the first route includes the estimated time to reach the road segments passed through in the first route, and the information of the second route includes the estimated time to reach the road segments passed through in the second route. Based on the first estimated time of arrival at the target road segment, the number of first vehicles arriving at the target road segment in each future time segment is determined; based on the second estimated time of arrival at the other road segments, the number of second vehicles arriving at the other road segments in each future time segment is determined; wherein, the first time distribution information is the number of first vehicles arriving at the target road segment in each future time segment, and the second time distribution information is the number of second vehicles arriving at the other road segments in each future time segment.

15. The apparatus according to any one of claims 12-14, wherein, The road segment travel time determination unit is specifically used for: The first road condition feature of the target road segment and the second road condition feature corresponding to the target road segment are input into a preset time prediction model to obtain the travel time of passing through the target road segment in each future time segment.

16. The apparatus according to any one of claims 12-14, The route travel time determination unit includes: The target sub-duration determination module is used to determine the target sub-duration based on the departure time corresponding to the target route planning request, from the travel sub-duration of the first target road segment that needs to be passed. The target sub-duration determination module is also used to determine the departure time of the next target road segment to be passed based on the departure time and the target sub-duration, and to determine the next target sub-duration in the passage sub-duration of the next target road segment; The total duration determination module is used to determine the total estimated travel time of the target route based on the target sub-duration of each target road segment.

17. The apparatus according to any one of claims 12-14, wherein, The target route planning request has multiple target routes; The device further includes a screening unit for: Based on the total estimated travel time of each target route, select the N routes with the shortest total estimated travel time from the target routes and return the N routes. N is an integer greater than or equal to 1.

18. A model training device for predicting the travel time of a road segment, comprising: A training data acquisition unit is used to acquire a training dataset, which includes a first road condition feature of a target road segment, a second road condition feature of other road segments connected to the target road segment, and a first travel time corresponding to the target road segment. The first and second road condition features are determined based on historical planned routes. The first road condition feature includes first distance distribution information and a first speed, and the second road condition feature includes second distance distribution information and a second speed. The first distance distribution information represents the distance between a first vehicle needing to enter the target road segment and the target road segment. The second distance distribution information represents the distance between a second vehicle needing to enter other road segments connected to the target road segment and the other road segments. The first speed is the speed of a vehicle located on the target road segment, and the second speed is the speed of a vehicle located on other road segments connected to the target road segment when the historical planned route was generated. The prediction unit is used to input the first road condition feature and the second road condition feature into a preset model to obtain the predicted travel time corresponding to the target road segment; the predicted travel time includes the travel sub-time of passing through the target road segment in each time segment after the preset time. An adjustment unit is used to adjust the parameters in the model according to the first travel duration and the predicted travel duration to obtain a model for estimating the travel duration of a road segment. The data preprocessing unit, including a distance information extraction module, is used for: Based on the information of the historical planned routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined in the historical planned routes; Obtain the first position of the first vehicle traveling along the first route at the preset time, determine the first distance between the first vehicle and the target road segment based on the first position, and determine the first distance distribution information based on each of the first distances; Based on the second route, determine the second position of the second vehicle that needs to pass through the other road segments at the preset time, and determine the second distance between the second vehicle and the target road segment based on the second position, and determine the second distance distribution information based on each of the second distances.

19. The apparatus according to claim 18, wherein the data preprocessing unit is further configured to: Obtain historical planned routes generated based on route planning requests prior to the preset time; Based on the information of the historical planned routes, determine the first time distribution information of the number of vehicles arriving at the target road segment after the preset time, and the second time distribution information of the number of vehicles arriving at the other road segments after the preset time; Based on historical logs, determine the first speed and the second speed; The first road condition feature further includes: first time distribution information; the second road condition feature further includes: second time distribution information.

20. The apparatus according to claim 19, wherein, The data preprocessing unit includes a time distribution information extraction module, used for: Based on the information of the historical planned routes, a first route including the target road segment and a second route including other road segments connected to the target road segment are determined in the historical planned routes; In the information of the first route, a first estimated time to reach the target road segment is obtained, and in the information of the second route, a second estimated time to reach the other road segments is obtained; wherein, the information of the first route includes the estimated time to reach the road segments passed through in the first route, and the information of the second route includes the estimated time to reach the road segments passed through in the second route. Based on the first estimated time of arrival at the target road segment, determine the number of first vehicles arriving at the target road segment in each time segment after the preset time; based on the second estimated time of arrival at the other road segments, determine the number of second vehicles arriving at the other road segments in each time segment after the preset time.

21. The apparatus according to any one of claims 19-20, wherein, The prediction unit is specifically used for: The first time distribution information, the first distance distribution information, and the first speed of the target road segment, as well as the second time distribution information, the second distance distribution information, and the second speed of the other road segments connected to the target road segment, are input into a preset model to obtain the predicted travel time through the target road segment in each time segment after the preset time.

22. The apparatus according to any one of claims 19-20, wherein the predicted travel time includes the travel sub-time of passing through the target road segment in each time segment; The data preprocessing unit further includes a supervisory data determination module, used for: The first passage duration for traveling the target road segment within the time segment is determined based on historical logs; the first passage duration includes the first passage duration for multiple time segments.

23. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-11.

24. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-11.

25. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1-11.

Images