Route recommendation method and apparatus

Abstract

Embodiments of the present application disclose a route recommendation method and device. Embodiments of the present application generate a plurality of candidate routes according to information of a plurality of ride orders obtained through matching, and then determine a route heat characteristic value and a bus lane characteristic value of each candidate route, and recommend one of the candidate routes as a target route based on the route heat characteristic value and the bus lane characteristic value. Since the bus lane characteristic is taken into account in the recommendation of the route, the route recommendation of a network bus service operated by a minibus or a bus and the like is more adapted to the scene requirements, the accuracy of the optimal route can be improved, and the operating cost is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data recommendation, and in particular to a route recommendation method and device, electronic equipment and a storage medium. BACKGROUND

[0002] In urban areas, the ride-hailing and carpooling business is developing rapidly. At the same time, users expect relevant travel service providers to provide more affordable services. More and more passengers expect to enjoy a travel service similar to a bus, with several stops along the way, but with more flexible routes and stops. In this mode of travel service, for vehicles with 7 seats or more participating in operation, how to reasonably plan the driving route is a problem to be solved. SUMMARY

[0003] Therefore, the embodiments of the present application provide a route recommendation method, device, electronic equipment and storage medium, so as to improve the recommendation accuracy of the optimal route and reduce the operating cost.

[0004] In a first aspect, the embodiments of the present application provide a route recommendation method, which comprises:

[0005] determining a plurality of pickup orders according to an order set, wherein the order set comprises a plurality of pickup requests;

[0006] generating a predetermined number of candidate routes according to the plurality of pickup orders;

[0007] determining a route heat feature value and a bus lane feature value of each candidate route, wherein the route heat value feature value is used to represent the number of historical orders covered by the candidate route, and the bus lane feature value is used to represent the proportion of the bus lane in the candidate route;

[0008] determining a first score of each candidate route according to the route heat feature value and the bus lane feature value;

[0009] determining the candidate route with the highest first score as a target route.

[0010] In a second aspect, the embodiments of the present application provide a route recommendation device, which comprises:

[0011] an order determination unit configured to determine a plurality of pickup orders according to an order set, wherein the order set comprises a plurality of pickup requests;

[0012] a candidate route determination unit configured to generate a predetermined number of candidate routes according to the plurality of pickup orders;

[0013] The characteristic value determining unit is configured to determine a route heat characteristic value and a bus lane characteristic value of each candidate route, the route heat characteristic value is used to represent a historical order quantity covered by the candidate route, and the bus lane characteristic value is used to represent a proportion of the bus lane in the candidate route.

[0014] The first score determining unit is configured to determine a first score of each candidate route according to the route heat characteristic value and the bus lane characteristic value.

[0015] The target route determining unit is configured to determine the candidate route with the highest first score as the target route.

[0016] In a third aspect, an embodiment of the present application provides a computer readable storage medium, which stores computer program instructions, and the computer program instructions are used to implement the method in the first aspect when executed by a processor.

[0017] In a fourth aspect, an embodiment of the present application provides an electronic device, which comprises a memory and a processor, and the memory is used to store one or more computer program instructions, and the one or more computer program instructions are executed by the processor to implement the method in the first aspect.

[0018] The technical scheme of the embodiment of the present application generates a plurality of candidate routes according to the information of the plurality of orders, and then determines the route heat characteristic value and the bus lane characteristic value of each candidate route, and recommends one of the candidate routes as the target route based on the route heat characteristic value and the bus lane characteristic value. Since the bus lane characteristic value is considered in the recommendation of the route, the route recommendation of the network bus service operated by a larger vehicle such as a minibus or a bus is more suitable for the scene requirement, the accuracy of the optimal route can be improved, and the operation cost can be reduced. BRIEF DESCRIPTION OF DRAWINGS

[0019] The above and other objects, features and advantages of the present application will become more apparent from the following description of embodiments of the present application, taken in conjunction with the accompanying drawings, in which:

[0020] Figure 1 is a system block diagram of the network bus service system of the embodiment of the present application.

[0021] Figure 2 is a flowchart of the route recommendation method of the embodiment of the present application;

[0022] Figure 3 is a schematic diagram of the network bus route of the embodiment of the present application;

[0023] Figure 4 is a schematic diagram of the network bus turning situation of the embodiment of the present application;

[0024] Figure 5 is another schematic diagram of a U-turn case of the online bus of an embodiment of the present application;

[0025] Figure 6 is a flowchart of determining a predetermined number of alternative routes of an embodiment of the present application;

[0026] Figure 7 is a flowchart of determining a set of alternative routes of an embodiment of the present application;

[0027] Figure 8 is a schematic diagram of a route recommendation device of an embodiment of the present application;

[0028] Figure 9 is a schematic diagram of an electronic device of an embodiment of the present application. DETAILED DESCRIPTION

[0029] The present application is described in the following based on embodiments, but the present application is not limited to these embodiments only. In the following detailed description of the present application, some specific details are described in detail. The present application can also be completely understood without the description of these details by those skilled in the art. In order to avoid obscuring the essence of the present application, well-known methods, processes, procedures, elements and circuits are not described in detail.

[0030] In addition, those of ordinary skill in the art should understand that the drawings provided herein are for illustrative purposes only and are not necessarily drawn to scale.

[0031] Unless the context clearly requires otherwise, throughout the description, the words "comprise", "comprising", and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is to say, in the sense of "including, but not limited to".

[0032] In the description of the present application, it should be understood that the terms "first", "second", and the like are used only for the purpose of description and should not be construed as indicating or implying relative importance. In addition, in the description of the present application, unless otherwise stated, the meaning of "multiple" is two or more.

[0033] Figure 1 is a system block diagram of the online bus service system of an embodiment of the present application. As shown in Figure 1 , the online bus service system of the present embodiment includes a server 1, a passenger end 2 and a driver end 3. The server 1, the passenger end 2 and the driver end 3 are connected to each other through a network 4, so as to realize the interaction of information and data. It should be understood that, although Figure 1 only a certain number of servers 1, passenger ends 2 and driver ends 3 are shown, this does not mean to limit the number of each, and multiple servers, passenger ends and driver ends can be included in the system.

[0034] The server 1 should be understood as a device providing data processing, database, communication facilities. For example, the server 1 can refer to a single physical server with associated communication and data storage and database facilities, or can refer to a collection of networked or clustered processors, associated networks and storage devices, and operating software and one or more database systems and application software supporting the services provided by the server. The server 1 can be a monolithic server or a distributed server across multiple computers or computer data centers, or can be various types of cloud servers. In some embodiments, each server can include hardware, software, or an embedded logic component or a combination of two or more such components for performing suitable functions supported or implemented by the server.

[0035] The passenger terminal 2 and the driver terminal 3 are communication terminals capable of running computer programs. These communication terminals can be mobile phones, tablet computers, palmtop computers, wearable devices, vehicle integrated terminals, etc. The passenger terminal 2 and the driver terminal 3 both have communication modules capable of wired or wireless communication. In some embodiments, the passenger terminal 2 and the driver terminal 3 include at least one remote communication module, such as a communication circuit for WLAN, GPRS, 2G / 3G / 4G / 5G remote communication. The passenger terminal 2 and the driver terminal 3 also have display devices and input devices. The display devices can be liquid crystal displays, LED displays or projection devices. The input devices can include touch screens, buttons, pressure sensors, etc. The passenger terminal 2 and the driver terminal 3 receive passenger instructions through the input devices and interact with the passengers or drivers through the display devices.

[0036] The passenger terminal 2 can run a passenger client 21. The driver terminal 3 can run a driver client 31. At the same time, the server 1 runs a server application 11. The server 1, the passenger terminal 2 and the driver terminal 3 can communicate through the network 4. In some application scenarios, the passenger can publish a ride request by operating the passenger client 21. The server 1 receives the ride request and matches the appropriate driver terminal 3 based on, for example, the location information reported by the driver terminal 3 and the location information of the ride request, and then sends the ride request to one or more matched driver terminals 3. Further, a ride order can be generated.

[0037] The web bus of the embodiment has certain differences from the existing web car service. The existing web car service usually does not set a fixed station, but the web car patrols in a certain area of the city, the server 1 matches the corresponding ride request according to the position of the web car, forms a ride order and sends the selected pickup location of the passenger to the driver end, and the driver end picks up and drops off the passenger at the selected pickup point and drop-off point according to the requirements of the ride task. In this business development process, the pickup point and the drop-off point are selected by the passenger through the passenger client, and can be any position that the web car can reach. Such service provides flexible personalized service as a taxi, and passengers can enjoy door-to-door travel service, but such web car service and taxi have high price. In order to reduce the price, carpooling service appears, that is, multiple passengers who do not know each other send carpooling service requests to the server, the server will match these carpooling service requests, and assign a specific web car driver end to multiple carpooling service requests. The driver end picks up and drops off multiple passengers according to the order of the carpooling service task to different positions. Carpooling service greatly reduces the service price and improves the utilization efficiency of the vehicle. However, the selection and determination of the pickup location and the drop-off location in the carpooling service are the same as the web car service, and are still selected by the passenger. The web car platform only recommends several pickup or drop-off positions with high frequency in the selected pickup location of the passenger, to assist the passenger in selection.

[0038] However, there is still a demand for further reducing the price of travel, and a demand for enjoying more flexible and convenient travel services than existing public transportation while having a lower price. Based on the above demand, there is a network bus service. The network bus has the concept of fixed stations and integrates the logic of existing carpooling services. However, unlike existing bus services, existing bus services are operated based on completely unchanged routes according to different bus lines. The network bus service forms a ride order based on the user's ride request, inheriting the mechanism of initiating online based on user requests in network car services and carpooling services. On the other hand, the network bus service limits the pickup and drop-off locations to fixed stations as much as possible. The advantage of setting fixed stations is that more passengers can be picked up and dropped off at a fixed station at one time with a higher probability, improving efficiency. At the same time, in the case of setting fixed stations, passengers can initiate a ride request by scanning a code directly at the fixed station in addition to initiating a request online. At the same time, the network bus service still has high flexibility, and the server can match the ride order according to the location of the passenger end initiating the request and the location of the driver end of the network bus in a carpooling-like logic. Moreover, the network bus is not running according to the set fixed route all day and all period, but running around the area covered by multiple different stations in the city area, retaining the flexibility of network car and taxi. Moreover, in some modes, the driving route of the network bus can be dynamically changed during driving, with high intelligence, which can meet the travel needs of passengers with high efficiency.

[0039] In Figure 1 In the network bus service system shown, in a typical example, the passenger end 2 can initiate a ride request to the server 1 through the passenger client 21, and the ride request can have the pickup station and the drop-off station (both are fixed stations) selected by the passenger. The server 1 can match and send the corresponding ride task to the driver end 3 of the network bus running to the vicinity of the drop-off station according to the network bus running around the pickup station, forming a ride order. The driver end 3 guides the network bus to drive to the pickup station through the driver client 31, and along the predetermined driving route, passes through multiple fixed stations, and sends the passenger to the drop-off station. In this process, the driver end 3 of the network bus is still in the order receiving state, and can continuously match new ride requests and pick up passengers at the corresponding pickup station and send them to the drop-off station. The server 1 can plan the route of the network bus in real time according to the actual distribution of the received ride requests, so that the network bus runs between different stations according to the planned route to pick up and drop off passengers.

[0040] In the business system of the online bus, when assigning a ride task to an empty online bus, a group of ride orders matched is usually assigned to the online bus, and a route for the online bus to complete the group of ride orders is planned. When the online bus is a larger vehicle, how to adapt to the business mode to dynamically recommend a suitable route passing through several stations for the vehicle becomes a problem to be solved.

[0041] Figure 2 is a flowchart of a route recommendation method of an embodiment of the present application, which is applied to a server. As shown in Figure 2 The route recommendation method comprises the following steps:

[0042] Step S100, a plurality of ride orders are determined according to an order set, and the order set comprises a plurality of ride requests.

[0043] The order set is a ride order sent by a passenger end and received by the server. In this step, a plurality of ride orders are determined in the order set. Specifically, a pre-trained order matching model can be used for ride order matching. Specifically, the order matching model is trained to match orders as much as possible to avoid U-turn routes. This is because the fewer U-turn routes, the higher the safety and stability of the online bus driving, and the time can be saved.

[0044] Specifically, the passenger initiates a ride request in the passenger end, and the ride request includes user information and ride information. The user information can be a user identifier of the passenger. The ride information can include start point information and end point information. The passenger end is a terminal device held by the passenger.

[0045] It should be noted that in actual application scenarios, in addition to matching a plurality of ride orders before departure, similar to carpooling of online car-hailing, the server can also assign newly received orders to the online bus. After determining to assign new orders to the online bus, the plurality of ride orders need to add new orders on the basis of the original. At this time, since the route has changed, it may be necessary to re-recommend the route.

[0046] It should be noted that when the driver end and the passenger end are relatively close, for example, the passenger end is within the preset range of the driver end, the order should be avoided to be assigned to the online bus to avoid traffic accidents caused by sudden braking.

[0047] Step S200, a predetermined number of candidate routes are generated according to the plurality of ride orders.

[0048] After determining the plurality of ride orders, it is necessary to generate candidate routes according to the ride orders, the candidate routes being routes passing through the starting points or ending points of the ride orders. For each candidate route, the plurality of passengers corresponding to the plurality of ride orders can realize boarding at the starting point and alighting at the ending point in the process of the network bus passing through the candidate route.

[0049] Due to the difference between the network bus and the ordinary car, in the process of generating the candidate route, the impassable road sections of the network bus need to be considered, and the candidate route is generated after removing these impassable road sections. Due to the large difference in volume between the network bus and the small passenger car with less than 7 seats, some low-value road sections with height limit and width limit are impassable for the network bus. In addition, due to the influence of traffic rules, there will also be some routes that the network bus cannot pass, such as a small passenger car access road. Therefore, in the scenario of the network bus, in the process of generating the candidate route, these impassable road sections can be removed from all routes of the road network first, and the candidate route is generated on the remaining road sections, or the route with the impassable road section is removed after the route is generated.

[0050] Figure 3 is a schematic diagram of the network bus route planning of an embodiment of the present application. As shown in Figure 3 , the route includes a plurality of stations A, B, C, and D. It should be noted that the stations can be bus stations or stations suitable for network bus operation mined by analyzing driving data of network bus services. It should be understood that there are a plurality of passing routes between different stations, and the present embodiment only schematically shows two routes between two stations. In this embodiment, a road section refers to a basic constituent unit in a road network, i.e., a road section between two nodes (intersections or crossroads). Each road section can represent a street, a part of a highway, a bridge, a tunnel, or a specific road entity. Each road section has its own attributes, and when planning a route from one location to another, the route can be determined according to the attributes of the road section. Specifically, Figure 3It is to be noted that the dashed lines and the solid lines in the figure represent road segments, and different forms of lines are only used to distinguish different road segments in the same route. Specifically, there are two routes 31 and 32 between the station A and the station B, wherein the route 31 includes the road segment 31a and the road segment 31b, the route 32 includes the road segment 32a and the road segment 32b, there are two routes 33 and 34 between the station B and the station C, wherein the route 33 includes the road segment 33a and the road segment 33b, the route 34 includes the road segment 34a, the road segment 34b and the road segment 34c. There are two routes 37 and 38 between the station C and the station D, wherein the route 37 includes the road segment 37a, the road segment 37b and the road segment 37c, the route 38 includes the road segment 38a, the road segment 38b and the road segment 38c. There are two routes 35 and 36 between the station A and the station D, wherein the route 35 includes the road segment 35a and the road segment 35b, the route 36 includes the road segment 36a and the road segment 36b. For example, when the starting point is the station A and the ending point is the station C, an optional route can be composed of the road segment 32a, the road segment 32b, the road segment 33a and the road segment 33b.

[0051] Specifically, when generating the alternative routes, at least one of the number of U-turn routes, the estimated time of the routes or the length of the routes is minimized, and a predetermined number of alternative routes passing through the starting point and the ending point of each of the plurality of ride orders are generated according to the set of impassable road segments and the plurality of ride orders.

[0052] Figure 4 is a schematic diagram of a U-turn situation of a web-bus according to an embodiment of the present application, as Figure 4 shown, the web-bus needs to make a U-turn at the road segment, and since the length of the web-bus is relatively large, the U-turn cannot be realized at the road segment. That is, although the relevant road segment is available for the web-bus to travel through, it is not available for the web-bus to make a U-turn. If the route contains such a road segment, and a U-turn route is planned at the location of the road segment, it can result in the web-bus being unable to travel according to the route or consuming a large amount of time for the U-turn.

[0053] Therefore, for a web-bus with a certain length and width, the minimum width required for the U-turn is determined, and the greater the width of the road segment during the U-turn, the higher the safety of the U-turn.

[0054] Figure 5 is another schematic diagram of a U-turn situation of a web-bus according to an embodiment of the present application, as Figure 5 shown, the road segment is wider than Figure 4 the road segment shown in Figure 4 , and therefore, the web-bus can make a U-turn at the road segment. Therefore, the route containing the U-turn at this location can be used as an alternative route. Conversely, the route containing the U-turn at the location shown in

[0055] Figure 6 is a flowchart of determining a predetermined number of alternative routes according to an embodiment of the present application, as shown in Figure 6 The determining a predetermined number of alternative routes includes the following steps:

[0056] Step S210, generating an alternative route set composed of alternative routes passing through the starting point and the ending point of each ride order and not containing the non-passable road segments in the non-passable road segment set.

[0057] In an optional implementation, after the non-passable road segments are determined, the road segments in the non-passable road segment set are removed from the total road segment set composed of all road segments, and the set composed of the obtained road segments is the passable road segment set. It should be noted that, since the network bus can travel on the bus lane compared with the ordinary car, the bus lane is added when determining the set composed of all road segments.

[0058] After the passable road segment set is determined, the alternative route set passing through the starting point and the ending point of each ride order is generated according to the passable road segment set. Specifically, the alternative route can be obtained by a pre-trained route generation model, wherein the route generation model is pre-trained to generate a route without non-passable road segments and / or a route without a U-turn position with a road width not meeting the predetermined requirement.

[0059] In addition to directly obtaining the route meeting the predetermined requirement by the pre-trained route generation model, the route meeting the requirement can also be screened from all routes, as shown in Figure 7 Figure 7 is a flowchart of determining the alternative route set according to an embodiment of the present application, and the determining the alternative route set can include the following steps:

[0060] Step S211, generating an initial alternative route set, wherein the initial alternative route set includes initial alternative routes passing through the starting point and the ending point of each ride order.

[0061] Specifically, all initial alternative routes including the starting point and the ending point of each ride order are generated. It should be noted that in the initial alternative route, it must be ensured that for any one ride order, the ending point is after the starting point.

[0062] Step S212, removing the route with non-passable road segments and / or the route with a U-turn position with a road width not meeting the predetermined requirement from the initial alternative route set, and determining the alternative route set.

[0063] ​Specifically, after all the initial optional routes are determined, it is necessary to remove the routes that do not meet the requirements from the initial optional routes. In the embodiment of the present application, when the initial optional routes include a route with an impassable road segment or a route with a U-turn position whose width does not meet the predetermined requirement, it is necessary to remove the route. Take the route shown in FIG. 3 as an example. If an optional route includes the route 33, and the road segment 33a in the route 33 is an impassable road segment, the initial optional route is removed from the initial optional route set. If an initial optional route needs to make a U-turn at a position of the road segment 36a in the route 36, and the width of the road segment 36a at the U-turn position is less than the minimum width required by the network bus to make a U-turn, the initial optional route is removed from the initial optional route set. Figure 3 Take the route shown in FIG. 3 as an example. If an optional route includes the route 33, and the road segment 33a in the route 33 is an impassable road segment, the initial optional route is removed from the initial optional route set. If an initial optional route needs to make a U-turn at a position of the road segment 36a in the route 36, and the width of the road segment 36a at the U-turn position is less than the minimum width required by the network bus to make a U-turn, the initial optional route is removed from the initial optional route set.

[0064] Through the step S211 and the step S212, the optional routes are obtained, so that the sample size is reduced, and the calculation of the second score in the subsequent step is facilitated.

[0065] The step S220 determines the second score of each optional route in the optional route set.

[0066] The second score is determined according to at least one of the number of U-turn routes, the route estimated time, or the route length. Specifically, the fewer the number of U-turn routes in the optional route, the shorter the route estimated time, the shorter the route length, the higher the driving efficiency in the route, and the higher the second score of the optional route.

[0067] In an optional implementation, the second score is determined according to the number of U-turn routes. Take a full score of 100 points as an example, and a deduction system with a lower limit of 0 points is described. Each U-turn route deducts a first predetermined score. In an optional implementation, the first predetermined score is 10 points. For example, when there are 5 U-turn routes in a route, the second score is deducted by 50 points, and the second score is 50 points. It should be noted that the deduction value of the U-turn route can not be a fixed value, and can be determined according to the width of the road segment where the U-turn route is located. When the width of the road segment is wider, the deducted score can be correspondingly reduced.

[0068] In an optional implementation, the second score is determined according to the route estimated time, and the longer the route estimated time, the lower the second score. In an optional implementation, when the route estimated time is less than the standard time, or the difference from the standard time is within a predetermined time period, the second score is 100 points, and the route estimated time exceeds the standard time by a predetermined time period, and a second predetermined score is deducted. It should be noted that, due to the different number and content of orders, the standard time needs to be determined according to the actual situation, and in an optional implementation, the estimated time of the shortest optional route in the optional route is taken as the standard time. Taking the predetermined time period as 5 minutes and the second predetermined score as 5 points as an example, when the route estimated time is 62 minutes, since the difference between the route estimated time and the standard time is within 5 minutes, the second score of the route is 100 points. When the route estimated time is 92 minutes, since it exceeds the standard time by 6 predetermined time periods, 30 points are deducted, and the second score is 70 points. In addition, the second score can also be determined according to the preset correspondence between the route estimated time and the second score, for example, when the route estimated time is below 40 minutes, the second score is 100 points, when the route estimated time is between 40-60 minutes, the second score is 80 points, when the route estimated time is between 60-90 minutes, the second score is 70 points, and when the route estimated time is above 90 minutes, the second score is 50 points. Thus, after the route estimated time of the optional route is determined, the second score of the optional route is determined according to the interval in which the route estimated time is located, for example, when the route estimated time is 75 minutes, the second score is determined to be 70 points.

[0069] In an optional implementation, the second score is determined according to the route length, and the longer the route length, the lower the second score. In an optional implementation, similar to the route estimated time, when the route length is less than the standard length, or the difference from the standard length is within a predetermined distance, the second score is 100 points, and the route length exceeds the standard time by a predetermined distance, and a third predetermined score is deducted. It should be noted that, due to the different number and content of orders, the standard length needs to be determined according to the actual situation, for example, the length of the shortest optional route in the optional route is taken as the standard length. Taking the standard length as 3km, the predetermined distance as 200m, and the third predetermined score as 5 points as an example, when the route length is 3.7km, since it exceeds the standard length by three predetermined distances, 15 points are deducted, and the second score is 85 points.

[0070] In an optional implementation, the second score is determined according to at least two criteria, and the optional route is scored according to the two criteria, and the final score is taken as the target score. Taking the case where the second score is determined according to the number of U-turn routes and the route length as an example, when the first predetermined score is 10, the third predetermined score is 5, the number of U-turn routes is 2, the predetermined distance is 200 m, the standard length is 2 km, and the standard route length is 2.5 km, 20 points are deducted according to the number of U-turn routes, and 10 points are deducted according to the standard route length, so the second score of the optional route is 70.

[0071] It should be noted that when the second score is determined according to at least two criteria, the deduction degree of each criterion needs to be adjusted according to the actual influence. For example, if the U-turn route has a greater impact on the actual traffic, when the second score is determined according to the number of U-turn routes, the route estimated time, and the route length, the first predetermined score is set to be higher, for example, the first predetermined score is set to 25, and the second predetermined score and the third predetermined score are set to 5, so as to increase the influence degree of the number of U-turn routes.

[0072] In addition, when the second score is determined according to at least two criteria, the scores of each criterion can also be scored respectively, and the lowest score / the highest score is taken as the final second score. For example, if the lowest score is taken as the final second score, the score according to the number of U-turn routes is 80, and the score according to the route estimated time is 75, the second score is determined to be 75. In this way, the characteristics of the route can be highlighted more clearly from the score.

[0073] In addition to the preset scoring criteria, the second score can also be determined by a pre-trained second score determination model. Specifically, the second score determination model is trained to have the ability to score according to at least one of the number of U-turn routes, the route estimated time, and the route length. In an optional implementation, the second score determination model is trained by a historical route traveled by the online taxi as a sample, the sample is specifically the frequency of the historical route being adopted, the frequency is specifically the frequency of being filtered as a candidate route when the route appears in the optional route, and the label is the number of U-turn routes, the route estimated time, and the route length. Scoring by the second score determination model can preliminarily filter a large number of optional routes to reduce the amount of subsequent first score calculation.

[0074] Step S230, determining a predetermined number of optional routes with the highest second score in the set of optional routes as the candidate routes.

[0075] Specifically, after the second scores of the optional routes in the optional route set are determined, a predetermined number of optional routes with high second scores are selected from the optional routes as candidate routes, so as to subsequently select an optimal route from the candidate routes. For example, the optional route set includes 10 optional routes, and the second scores of the 10 optional routes are 95, 90, 85, 80, 80, 75, 60, 50, 20, and 20 respectively. Assuming that the predetermined number is 3, the candidate routes are the routes with second scores of 95, 90, and 85.

[0076] It should be noted that in the embodiments of the present application, the candidate routes with high second scores do not necessarily need to be selected according to the predetermined number. In an optional implementation, a criterion can be set to filter out some routes with too low scores. For example, the scores of 10 optional routes are 95, 80, 75, 50, 40, 45, 30, 20, 20, and 10 respectively. Assuming that the predetermined number is 5, the candidate routes are the routes with second scores of 95, 80, 75, 50, and 40. However, the routes with scores of 50 and 40 have too low scores and poor quality, and therefore, the two routes can be filtered out by setting a criterion that the second score is at least 60. The criterion is set according to actual conditions. In this way, some routes with poor passing quality can be filtered out, so as to reduce the calculation amount of the route heat feature value and the bus lane feature value in the subsequent calculation.

[0077] Through steps S210 to S230, the candidate routes with high second scores corresponding to the plurality of ride orders are determined, so as to reduce the number of routes that need to be calculated subsequently, and thus the calculation of the first score is facilitated.

[0078] In step S300, the route heat feature value and the bus lane feature value of each candidate route are determined. The route heat value feature value is used to represent the number of historical orders covered by the candidate route, and the bus lane feature value is used to represent the proportion of the bus lane in the candidate route.

[0079] The route heat feature value is used to represent the number of historical orders covered by the candidate route. Specifically, when all or part of the candidate route covers the travel path of a historical order, the route heat feature value of the candidate route is improved. Therefore, the higher the route heat feature value of a candidate route is, the more times the station in the candidate route has been used in history, which indicates that the route is more suitable to be used.

[0080] It should be noted that the historical orders covered by the candidate route are the same historical orders as the candidate route, and can also be historical orders contained in the candidate route, that is, the route corresponding to the historical order can be a sub-route of the candidate route. For example, the historical orders covered by the candidate route are the historical orders corresponding to the sub-routes of the candidate route. Figure 3The shown route is taken as an example for illustration. If the historical order corresponds to the route 31 and the route 34, and the alternative route includes the route 31, the route 34 and the route 37, the historical order will increase the heat of the alternative route. The historical order and the alternative route have overlapping parts, and the historical order will also increase the heat of the alternative route. For example, the historical order corresponds to the route 31, the route 34 and the route 37, and the alternative route includes the route 34, the route 37 and the route 35. The historical order will also increase the heat of the alternative route.

[0081] Specifically, the route heat can be calculated according to the degree of overlap between the historical order and the alternative route. In an optional implementation, the heat of the alternative route is increased according to the proportion of the overlapping part of the historical order in the alternative route. For example, the overlapping part of a certain historical order accounts for 30% of the alternative route, and the historical order of the same alternative route increases the heat of the alternative route by 100. Therefore, the historical order increases the heat of the alternative route by 30.

[0082] The bus lane feature value is used to represent the proportion of the bus lane in the alternative route. The higher the proportion of the bus lane, the more suitable the route is for the network bus. In an optional implementation, the bus lane feature value is proportional to the proportion of the bus lane in length. Taking the bus lane feature value between 0 and 1 as an example, specifically, when the bus lane feature value is 0, it means that there is no bus lane in the alternative route. When the bus lane feature value is 1, it means that all the road sections in the alternative route are bus lanes. When the alternative route has both bus lanes and non-bus lanes, the ratio of the length of the bus lane to the length of the alternative route is taken as the bus lane feature value. When the length of the alternative route is 5km, and the length of the bus lane is 4km, the bus lane feature value is 0.8. In addition to determining the bus lane feature value according to the proportion of the bus lane in length, the bus lane feature value can also be determined by the proportion of the driving time of the bus lane in the estimated time.

[0083] Step S400, determining the first score of each alternative route according to the route heat feature value and the bus lane feature value.

[0084] Specifically, after the predetermined number of alternative routes are obtained by the second score screening, the alternative routes need to be further screened with the route heat and the bus lane as screening conditions, which are more related to the network bus business, to obtain the target route that is more suitable for the network bus driving.

[0085] After the route heat characteristic value and the bus lane characteristic value are calculated, a first score needs to be determined according to the route heat characteristic value and the bus lane characteristic value, and the first score represents the advantages and disadvantages of the candidate route. In some embodiments, the first score is positively correlated with the route heat characteristic value, and the first score is positively correlated with the bus lane characteristic value. That is, the greater the route heat characteristic value and the bus lane characteristic value, the lower the first score.

[0086] In an alternative implementation, the first score is obtained by deducting the route heat characteristic value and the bus lane twice. Taking a full score of 100 points as an example, a deduction system with a lower limit of 0 points is described. Specifically, the score is determined according to the pre-set relationship between the route heat characteristic value / bus lane and the deduction. For example, when the route heat characteristic value is 0.8 or more, no deduction is made, when the route heat characteristic value is 0.6-0.8, 10 points are deducted, when the route heat characteristic value is 0.5-0.6, 30 points are deducted, and when the route heat characteristic value is 0.5 or less, 50 points are deducted; when the bus lane characteristic value is 0.6 or more, no deduction is made, when the bus lane characteristic value is 0.4-0.6, 10 points are deducted, and when the bus lane characteristic value is 0.4 or less, 30 points are deducted. At this time, if the route heat characteristic value of a certain candidate route is 0.7 and the bus lane characteristic value is 0.5, the first score of the candidate route is 80 points.

[0087] In another alternative implementation, the first score is obtained by weighted sum of the score corresponding to the route heat characteristic value and the score corresponding to the bus lane characteristic value. Taking a full score of 100 points as an example, assuming that the weight of the route heat characteristic value is 0.2 and the weight of the bus lane characteristic value is 0.8. First, the score corresponding to the route heat characteristic value and the score corresponding to the bus lane characteristic value are determined according to the route heat characteristic value and the bus lane characteristic value, and then the two scores are weighted. For example, the score corresponding to the route heat characteristic value of a certain candidate route is 70 points, and the score corresponding to the bus lane characteristic value is 90 points, and the weighted score is 86 points.

[0088] In yet another optional implementation, the first score can also be obtained by a pre-trained first score determination model. Specifically, the first score determination model is trained to have the ability to comprehensively score the routes by combining the bus lane feature value and the route heat feature value. In some embodiments, in addition to the route heat feature and the bus lane feature, the bus lane estimated time can also affect the first score. Specifically, the estimated time of the alternative route with the bus lane is calculated by a pre-trained time prediction model, and then the first score of each alternative route is determined according to the route heat feature, the bus lane feature, and the bus lane estimated time. Specifically, the longer the bus lane estimated time, the higher the first score. Specifically, since the traffic rules and road conditions of the bus lane are different from those of the ordinary road section, for example, the bus lane only allows buses to travel and does not allow small passenger cars to pass, the traffic speed of the bus lane is relatively stable, and in addition, the speed limit of the bus lane is also different from that of the ordinary lane. Therefore, the driving conditions of the bus lane are completely different from those of the ordinary lane, and the prediction by the ordinary time prediction model is not accurate enough, and needs to be predicted by a pre-trained time prediction model. The time prediction model is trained to be able to determine the time for passing each road section according to the type to which each road section belongs, wherein the type to which each road section belongs includes a bus lane and a non-bus lane.

[0089] Specifically, the time prediction model is trained by samples with bus lane passing time, and the samples are historical routes and the labels are times. The pre-trained time prediction model is used to calculate the estimated time, which can more accurately predict the time.

[0090] In the route recommendation method of the embodiments of the present application, two scores are involved, specifically, the second score is used to preliminarily screen all the alternative routes that meet the requirements to obtain a predetermined number of alternative routes, and then the first score determined according to the bus lane feature value and the route heat feature value is used to select the optimal route from the predetermined number of alternative routes. The purpose of screening by the second score is to preliminarily screen to reduce the amount of calculation when calculating the first score subsequently. The bus lane feature value is included in the calculation of the first score because the route with a high proportion of bus lanes only allows buses to travel, so it is more open and better for the driving conditions of the network bus.

[0091] In an optional implementation, the predetermined number of alternative routes corresponding to the second score can be obtained by reusing the existing route scoring model for network car hailing, and then the bus lane feature value and the route heat feature value are used to further select from the predetermined number of alternative routes that have been preliminarily screened.

[0092] Step S500, determining the candidate route with the highest first score as the target route.

[0093] When the first scores of the candidate routes are determined, the candidate route with the highest first score is determined as the target route.

[0094] The embodiment of the application generates a plurality of candidate routes according to the information of the matched plurality of ride orders, and further determines the route heat characteristic value and the bus lane characteristic value of each candidate route, and recommends one of the candidate routes as the target route based on the route heat characteristic value and the bus lane characteristic value. Since the bus lane characteristic is taken into account in the recommendation of the route, the route recommendation of the online bus service operated by larger vehicles such as minibuses or buses is more adapted to the scene demand, which can improve the accuracy of the optimal route and reduce the operating cost.

[0095] Figure 8 is a schematic diagram of a route recommendation device according to an embodiment of the application, as shown in Figure 8 The route recommendation device includes an order determination unit 81, a candidate route determination unit 82, a characteristic value determination unit 83, a first score determination unit 84, and a target route determination unit 85. The order determination unit 81 is configured to determine a plurality of ride orders according to an order set, and the order set includes a plurality of ride requests. The candidate route determination unit 82 is configured to generate a predetermined number of candidate routes according to the plurality of ride orders. The characteristic value determination unit 83 is configured to determine the route heat characteristic value and the bus lane characteristic value of each candidate route, the route heat value characteristic value is used to represent the number of historical orders covered by the candidate route, and the bus lane characteristic value is used to represent the proportion of the bus lane in the candidate route. The first score determination unit 84 is configured to determine the first score of each candidate route according to the route heat characteristic value and the bus lane characteristic value. The target route determination unit 85 is configured to determine the candidate route with the highest first score as the target route.

[0096] The embodiment of the application generates a plurality of candidate routes according to the information of the matched plurality of ride orders, and further determines the route heat characteristic value and the bus lane characteristic value of each candidate route, and recommends one of the candidate routes as the target route based on the route heat characteristic value and the bus lane characteristic value. Since the bus lane characteristic is taken into account in the recommendation of the route, the route recommendation of the online bus service operated by larger vehicles such as minibuses or buses is more adapted to the scene demand, which can improve the accuracy of the optimal route and reduce the operating cost.

[0097] Figure 9 is a schematic diagram of an electronic device according to an embodiment of the application. As shown in Figure 9 Figure 9 ​The electronic device shown is a general purpose data processing device that includes a general purpose computer hardware structure that includes at least a processor 91 and a memory 92. The processor 91 and the memory 92 are connected by a bus 93. The memory 92 is adapted to store instructions or programs executable by the processor 91. The processor 91 can be a single independent microprocessor or a collection of one or more microprocessors. In this manner, the processor 91 performs the processes and controls the other devices by executing the instructions stored in the memory 92 to implement the method flow of the embodiments of the present application as described above. The bus 93 connects the above components together and also connects the above components to a display controller 94 and a display device and input / output (I / O) devices 95. The input / output (I / O) devices 95 can be a mouse, a keyboard, a modem, a network interface, a touch input device, a body sensor input device, a printer, and other devices known in the art. Typically, the input / output devices 95 are connected to the system through an input / output (I / O) controller 96.

[0098] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, apparatus (devices) or computer program products. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer readable storage media (including, but not limited to, disk memory, CD-ROMs, optical storage devices, etc.) embodying computer readable program code.

[0099] The present application is described with reference to flow diagrams of methods, apparatus (devices) and computer program products according to embodiments of the present application. It will be understood that each flow diagram block, and combinations thereof, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processing device, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flow diagram block or blocks.

[0100] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flow diagram block or blocks. Figure 1 The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operations to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus implement the functions specified in the flow diagram block or blocks.

[0101] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flow diagram block or blocks. Figure 1 The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operations to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus implement the functions specified in the flow diagram block or blocks.

[0102] Another embodiment of the present application relates to a non-volatile storage medium for storing a computer readable program for causing a computer to execute some or all of the above-mentioned method embodiments.

[0103] That is, those skilled in the art can understand that all or part of the steps in the above-mentioned method embodiments can be completed by a program stored in a storage medium, including a plurality of instructions for causing a device (which can be a single-chip microcomputer, a chip, etc.) or a processor to execute all or part of the steps of the method described in the embodiments of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various storage media that can store program codes.

[0104] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A route recommendation method, characterized in that, The method includes: Multiple ride orders are determined based on an order set by a pre-trained order matching model, the order set including multiple ride requests, wherein the order matching model is trained to match orders with the criterion of avoiding U-turn routes as much as possible; Generate a predetermined number of alternative routes based on multiple ride orders; Determine the route popularity feature value and bus lane feature value for each candidate route. The route popularity feature value is used to characterize the number of historical orders covered by the candidate route, and the bus lane feature value is used to characterize the proportion of bus lanes in the candidate routes. The first score of each candidate route is determined based on the route popularity characteristic value and the bus lane characteristic value. The candidate route with the highest first-place rating is selected as the target route. Specifically, determining the first score for each candidate route based on the route popularity feature value and the bus lane feature value involves: The travel time of each alternative route is determined based on a time prediction model, wherein the time prediction model is trained based on historical data of travel time with bus lanes. The first score for each candidate route is determined based on the route popularity characteristics, bus lane characteristics, and estimated bus lane travel time. The characteristic value of the dedicated bus lane is determined based on the ratio of the length of the dedicated bus lane to the length of the alternative routes. Among them, the number of alternative routes generated based on multiple ride orders includes: Determine a set of impassable road sections, which are road sections where buses are not allowed to pass; The alternative routes are determined based on the set of impassable road sections and the multiple ride orders; The selection of alternative routes based on the set of impassable road sections and the multiple ride orders includes: With the objective of minimizing at least one of the following: the number of U-turn routes, the estimated route time, or the route length, a predetermined number of alternative routes passing through the origin and destination of each ride order are generated based on the set of impassable road sections and the multiple ride orders.

2. The method according to claim 1, characterized in that, With the objective of minimizing at least one of the following: the number of U-turn routes, estimated route time, or route length, a predetermined number of alternative routes passing through the origin and destination of the multiple ride orders are generated based on the set of impassable road segments and the multiple ride orders. Generate a set of optional routes that pass through the origin and destination of each ride order and do not include the impassable road segments in the set of impassable road segments; Determine the second rating for each optional route in the set of optional routes; A predetermined number of optional routes with the highest second rating from the set of optional routes are selected as candidate routes; The second score is determined based on at least one of the following: the number of U-turn routes, the estimated route time, or the route length.

3. The method according to claim 2, characterized in that, The set of optional routes, which generates routes passing through the origin and destination of each passenger order and does not include impassable road segments from the set of impassable road segments, includes: Generate an initial set of optional routes, which includes initial optional routes that pass through the origin and destination of each ride order; The optional route set is determined by removing routes with impassable sections and / or routes with U-turns where the road width at the U-turn location does not meet the predetermined requirements from the initial set of optional routes.

4. The method according to claim 2, characterized in that, The set of optional routes, which consists of optional routes that pass through the origin and destination of each ride order and do not include impassable road segments from the set of impassable road segments, includes: A set of optional routes is determined by a pre-trained route generation model, wherein the route generation model is pre-trained to generate routes that do not have impassable sections and / or do not have U-turn widths that do not meet predetermined requirements.

5. The method according to claim 1, characterized in that, The first score is positively correlated with the route popularity feature value, and the first score is positively correlated with the bus lane feature value.

6. A route recommendation device, characterized in that, The device includes: An order confirmation unit is used to determine multiple ride orders based on an order set using a pre-trained order matching model. The order set includes multiple ride requests. The order matching model is trained to match orders with the criterion of minimizing U-turn routes. The alternative route determination unit is used to generate a predetermined number of alternative routes based on multiple ride orders; The feature value determination unit is used to determine the route popularity feature value and bus lane feature value of each candidate route. The route popularity feature value is used to characterize the number of historical orders covered by the candidate route, and the bus lane feature value is used to characterize the proportion of bus lanes in the candidate routes. The first scoring determination unit is used to determine the first score of each candidate route based on the route popularity characteristic value and the bus lane characteristic value. The target route determination unit is used to determine the candidate route with the highest first score as the target route. Specifically, the first scoring determination unit is used for: The travel time of each alternative route is determined based on a time prediction model, wherein the time prediction model is trained based on historical data of travel time with bus lanes. The first score for each candidate route is determined based on the route popularity characteristics, bus lane characteristics, and estimated bus lane travel time. The characteristic value of the dedicated bus lane is determined based on the ratio of the length of the dedicated bus lane to the length of the alternative routes. The alternative route determination unit is used for: Determine a set of impassable road sections, which are road sections where buses are not allowed to pass; The alternative routes are determined based on the set of impassable road sections and the multiple ride orders; The selection of alternative routes based on the set of impassable road sections and the multiple ride orders includes: With the objective of minimizing at least one of the following: the number of U-turn routes, the estimated route time, or the route length, a predetermined number of alternative routes passing through the origin and destination of each ride order are generated based on the set of impassable road sections and the multiple ride orders.

7. A computer-readable storage medium storing computer program instructions thereon, characterized in that, The computer program instructions, when executed by a processor, implement the method as described in any one of claims 1-5.

8. An electronic device comprising a memory and a processor, characterized in that, The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method as described in any one of claims 1-5.

Images