Methods, apparatus, equipment, media, and procedures for vehicle dispatching recommendations.
By combining historical experience and vehicle status information with interaction methods and devices on terminal devices and servers, the problem of mismatch between recommended locations and execution costs in vehicle dispatching recommendations is solved, achieving consistency between interface status and underlying settings and ensuring the executability of recommendations.
Patent Information
- Application Number
- CN202611559928.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-09
- Publication Date
- 2026-11-17
AI Technical Summary
In vehicle dispatch recommendation, existing technologies suffer from a mismatch between recommended locations and actual vehicle execution costs, resulting in inconsistencies between the interface state and the underlying execution results, and a lack of state correlation between events and execution receipt verification.
By implementing interaction methods and devices on the terminal device and the server respectively, location information and setting suggestions are presented, confirmed, adjusted and processed. Combining historical experience and vehicle status information, the recommended location and order acceptance settings are determined, and the unified candidate structure and executability are verified to ensure the consistency between the interface state and the underlying setting state.
This reduces inconsistencies between the interface state and the underlying settings, improves the executability of recommendations, and prevents invalid navigation operations through deterministic checks, ensuring the accuracy and consistency of recommendations.
Smart Images

Figure CN123115531A_ABST
Abstract
Description
Technical Field
[0001] Examples of this disclosure relate to the fields of computer data processing and intelligent transportation, and more particularly to a method, apparatus, device, medium, and program product for vehicle scheduling recommendation. Background Technology
[0002] With the development of mobile communication and intelligent connected vehicle technologies, vehicle dispatch recommendations are widely used in transportation service scenarios such as ride-hailing and mobility platforms. During vehicle patrol or waiting for orders, the platform needs to provide vehicles with actionable dispatch recommendations to assist drivers in rationally planning their routes and adjusting their order-accepting behavior.
[0003] In some vehicle dispatch recommendation implementations, the data sources and processing granularity upon which the recommendations are based may differ. For example, historical order statistics, regional supply and demand status, and individual vehicle displacement costs may not be processed within the same evaluation framework, thus the recommended location may not match the actual execution cost of a specific vehicle. On the other hand, location recommendation, order setting confirmation, vehicle arrival, and underlying setting execution are different events; if there is a lack of state correlation between events and execution receipt verification, the setting status displayed on the interface may be inconsistent with the underlying execution results.
[0004] Therefore, it is necessary to provide a vehicle scheduling recommendation scheme that combines historical experience, current status, and vehicle displacement costs to improve the feasibility of the recommendation and ensure consistency between the confirmation of the setting suggestion, execution upon arrival, and the presentation of the interface results. Summary of the Invention
[0005] In a first aspect, this paper provides an interactive method for vehicle dispatch recommendation, executed by a terminal device. The method includes: presenting location information and setting suggestions in a cruise interface, wherein the location information indicates a recommended location, and the setting suggestions indicate a recommended status for order acceptance settings, the recommended status being associated with the recommended location; receiving a confirmation instruction for the setting suggestions; based on the confirmation instruction, triggering an adjustment of the order acceptance settings to the recommended status in response to the vehicle arriving at the recommended location; and in response to receiving an execution receipt, presenting a successful adjustment prompt in the cruise interface, the execution receipt indicating that the order acceptance settings have been adjusted to the recommended status.
[0006] In a second aspect of this paper, a vehicle dispatch recommendation method is provided, executed by a server. This method includes: acquiring experience records and vehicle status information, where the experience records are statistically derived from historical order-taking data and indicate applicable conditions, as well as the corresponding order-taking locations and their status; determining candidate locations based at least on the matching results of the status information and the applicable conditions, where the candidate locations satisfy execution constraints for the vehicle; determining recommended locations and recommended status for order-taking settings based on the matching results and the net value of the candidate locations, where the recommended locations are selected from the candidate locations, the recommended status is determined based on the recommended locations and the order-taking status indicated by the experience records, and the net value is determined based on the expected revenue and displacement costs determined for the vehicle and the corresponding candidate locations; and providing location information and setting suggestions to terminal devices, where the location information indicates the recommended location and the setting suggestions indicate the recommended status.
[0007] In a third aspect of this document, an interactive device for vehicle dispatch recommendation is provided, installed on a terminal device. The interactive device includes: a first presentation module configured to present location information indicating a recommended location and a setting suggestion indicating a recommended order-taking setting status on a cruise interface, the recommended status being associated with the recommended location; a receiving module configured to receive a confirmation instruction for the setting suggestion; a triggering module configured to, based on the confirmation instruction, trigger an adjustment of the order-taking setting to the recommended status when the vehicle arrives at the recommended location; and a second presentation module configured to, upon receiving an execution receipt indicating that the order-taking setting has been adjusted to the recommended status, present an adjustment success prompt on the cruise interface.
[0008] In a fourth aspect of this paper, a vehicle dispatch recommendation device is provided, located on a server. The device includes: an acquisition module configured to acquire experience records and vehicle status information based on historical order-taking data statistics, wherein the experience records indicate applicable conditions and the corresponding order-taking locations and order-taking settings status; a candidate determination module configured to determine candidate locations that satisfy execution constraints for vehicles, based at least on the matching results of status information and applicable conditions; a recommendation determination module configured to determine recommended locations and recommended order-taking settings status based on the matching results and the net value of candidate locations, wherein recommended locations are selected from candidate locations, the recommended status is determined based on the recommended locations and the order-taking setting status indicated by the experience records, and the net value is determined based on the expected revenue and displacement costs determined for the vehicle and the corresponding candidate locations; and a provision module configured to provide location information and setting suggestions to terminal devices.
[0009] In a fifth aspect of this document, an electronic device is provided. The electronic device includes at least one processor and at least one memory coupled to the at least one processor, the at least one memory storing instructions executable by the at least one processor, which, when executed, cause the electronic device to perform the methods of the first or second aspect.
[0010] In a sixth aspect of this document, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program that can be executed by a processor to implement the methods of the first or second aspect.
[0011] In a seventh aspect of this document, a computer program product is provided. This computer program product is tangibly stored in a non-transitory computer-readable storage medium and includes computer-executable instructions that, when executed by a processor, cause the processor to perform the method of the first or second aspect.
[0012] In this way, on the terminal interaction side, the confirmation status of the suggested settings, the vehicle arrival event, the actual execution status of the order settings, and the interface presentation status are recorded separately. The actual execution receipt returned by the settings interface is used as the basis for presenting the successful adjustment prompt, thereby reducing inconsistencies between the interface status and the underlying settings status, as well as repeated calls to the settings interface. On the server side, by adopting a unified candidate structure for candidate locations from different sources and performing integrity and executability checks before recommendation, invalid navigation operations caused by inconsistencies in location names, coordinates, or data validity periods can be reduced. By performing deterministic verification of the language model output, setting suggestions for locations outside the whitelist, locations with inconsistent coordinates, or locations without operation permissions can be blocked from being issued.
[0013] It should be understood that the description in the Summary of the Invention section is not intended to limit the key or essential features of the examples in this disclosure, nor is it intended to restrict the scope of the solutions. Other features will become readily apparent from the following description. Attached Figure Description
[0014] The above and other features, advantages, and aspects of the various examples in this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein: Figure 1 A schematic diagram of an example implementation environment of a vehicle dispatch recommendation system according to some embodiments of the present disclosure is shown; Figure 2 A flowchart illustrating an example process of vehicle dispatch recommendation interaction performed by a terminal device according to some embodiments of the present disclosure is shown; Figure 3 A schematic diagram of an example interface for the cruise function entry according to some embodiments of the present disclosure is shown; Figure 4 A schematic diagram of an example interface for receiving order preference input according to some embodiments of the present disclosure is shown; Figure 5A schematic diagram of an example cruise interface that first presents recommended locations and associated settings suggestions according to some embodiments of the present disclosure is shown; Figure 6 A schematic diagram of an example cruise interface for listening to a tour in place when the recommended location is the vehicle's current location, according to some embodiments of the present disclosure; Figure 7 A schematic diagram of an example cruise interface is shown, illustrating setup suggestions according to some embodiments of the present disclosure, where the recommended location has been confirmed but has not yet been reached. Figure 8 A schematic diagram of an example cruise interface is shown, illustrating how navigation continues to a recommended location when setting suggestions are not adopted, according to some embodiments of this disclosure; Figure 9 A schematic diagram of an example cruise interface showing navigation status to a recommended location according to some embodiments of the present disclosure is shown; Figure 10 A schematic diagram of an example cruise interface for yaw correction and maintaining a recommended location according to some embodiments of the present disclosure is shown; Figure 11 A schematic diagram of an example cruise interface for continuous navigation that maintains a valid recommended target according to some embodiments of the present disclosure is shown; Figure 12 A schematic diagram of an example cruise interface is shown during the settings adjustment process after arriving at a recommended location, according to some embodiments of the present disclosure; Figure 13 A schematic diagram is shown illustrating an example cruise interface that presents a successful settings adjustment after receiving an execution receipt according to some embodiments of the present disclosure; Figure 14 A schematic diagram of an example cruise interface showing a single prompt after reaching a first duration threshold according to some embodiments of the present disclosure is shown; Figure 15 A schematic diagram of an example cruise interface is shown, which updates recommended locations and setting suggestions after a second duration threshold is reached, according to some embodiments of the present disclosure. Figure 16 A schematic diagram of an example interface for proceeding to the pick-up point after receiving an order, according to some embodiments of the present disclosure, is shown; Figure 17 A schematic diagram of an example interface for trip completion and order feedback according to some embodiments of the present disclosure is shown; Figure 18 A schematic diagram of an example interface for feedback on the receipt of related rights according to some embodiments of this disclosure is shown; Figure 19 A flowchart illustrating an example process of vehicle dispatch recommendation performed by a server according to some embodiments of the present disclosure is shown; Figure 20A schematic diagram illustrating an example cruise interface showing a state description when no executable proposal is obtained according to some embodiments of this disclosure; and Figure 21 A block diagram of an electronic device capable of implementing embodiments of the present disclosure is shown. Detailed Implementation
[0015] Examples of this disclosure will now be described in more detail with reference to the accompanying drawings. While some examples of this disclosure are shown in the drawings, it should be understood that solutions can be implemented in various forms and should not be construed as limited to the examples set forth herein. Rather, these examples are provided to provide a more thorough and complete understanding of the solutions. It should be understood that the drawings and examples in this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of the solutions.
[0016] It should be noted that the headings of any section / subsection provided in this disclosure are not restrictive. Various examples are described throughout this document, and examples of any type may be included under any section / subsection. Furthermore, examples described in any section / subsection may be combined in any way with any other examples described in the same section / subsection and / or different sections / subsections.
[0017] In the description of examples in this disclosure, the term "including" and similar terms should be understood as open inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The terms "an example" or "the example" should be understood as "at least one example". The term "some examples" should be understood as "at least some examples". The terms "first", "second", etc., are used to distinguish different objects, and are not necessarily a restriction on order, importance, or size unless explicitly defined in the context; the "first duration threshold" and "second duration threshold" involved in this disclosure have a clear progressive relationship of time length, i.e., the second duration threshold is greater than the first duration threshold.
[0018] The examples in this document may involve user data, data acquisition, and / or use. All of these aspects comply with relevant laws, regulations, and provisions. In the examples, all data collection, acquisition, processing, forwarding, and use are conducted with the user's knowledge and confirmation. Accordingly, when implementing each example, the type, scope of use, and usage scenarios of any data or information involved should be communicated to the user and their authorization obtained through appropriate means in accordance with relevant laws and regulations. The specific methods of notification and / or authorization may vary depending on the actual situation and application scenario, and the scope of the solution is not limited in this regard. If the solutions in this specification and its examples involve the processing of personal information, such processing will be conducted on the premise of a legal basis (e.g., obtaining the consent of the personal information subject, or being necessary for the performance of a contract), and will only be conducted within the scope stipulated or agreed upon. A user's refusal to process personal information beyond what is necessary for basic functions will not affect the user's use of basic functions.
[0019] In some embodiments, historical data used to form experience records have identifying information that can directly identify a specific vehicle operator removed before statistical processing. Experience records are formed using statistical dimensions such as city, time period, service evaluation parameter range, and order acceptance method category, and do not store or output the identity information, complete historical location trajectory, or original income details of a specific vehicle operator. The vehicle's current location and current setting status are used for vehicle dispatch recommendations in this round and are used according to the data range required for recommendation processing. Statistical results formed from multiple historical samples are used to determine candidate locations or setting suggestions, and are not used to evaluate the qualifications or rights of a specific vehicle operator.
[0020] The term "experience record" as used in this article refers to a structured data asset obtained from historical order data, indicating applicable conditions and the corresponding order locations and order setting status. Experience records can be constructed separately by city and service evaluation parameters, and can be reconstructed using new historical data or updated periodically according to a preset update cycle.
[0021] The term "applicable conditions" as used in this document refers to the matching conditions that must be met for an experience record to be applicable to the current vehicle status. Applicable conditions may include at least one of the following: spatiotemporal conditions such as city, date type, and time period; the numerical range of the service evaluation parameters; the order acceptance method category determined based on historical order acceptance behavior statistics; and historical waiting time statistics. The numerical range of the service evaluation parameters and the order acceptance method category are used to retrieve experience records formed under the corresponding statistical conditions and do not represent a subjective value evaluation of a specific vehicle operator.
[0022] The term "order pickup location" as used in this document generally refers to a physical geographical location or area where vehicles can park, cruise, or wait to receive orders. In one or more specific embodiments, the order pickup location can be formed by quantizing and spatially aggregating the starting locations of historical orders, and can be associated with display names, aggregated representative coordinates, navigation coordinates, order tag distribution, and aggregated quality statistics. The aggregated representative coordinates are used to characterize the spatial center of historical order pickup locations; the navigation coordinates are coordinates associated with the order pickup location and verified through road accessibility, used for map marking, navigation endpoints, and arrival determination. The server uses the same location identifier to associate the display name, aggregated representative coordinates, and navigation coordinates to maintain the correspondence between the recommendation description, map markings, navigation, and arrival events.
[0023] The term "order acceptance settings" as used in this document refers to one or more business switches in a client or service platform used to control the permission to receive specific types of orders. For example, order acceptance settings may include preferential order acceptance switches, carpooling order acceptance switches, flexible order acceptance switches, long-distance order switches, and terminal order switches. The recommended state of the order acceptance settings refers to the state (e.g., on or off) suggested for the corresponding switch after the vehicle arrives at the recommended location. The actual adjustment of the setting state is executed by the underlying controlled deterministic setting interface.
[0024] The term "candidate whitelist" as used in this article refers to a set of candidate locations that have passed data integrity and executability checks. Executability checks may involve at least one of the following: physical road accessibility, urban traffic restrictions, maximum permissible displacement distance, data validity period, preset security restrictions related to continuous vehicle operation time, reservation status, and relevant operational permissions.
[0025] As used herein, the term "net value" refers to an evaluation metric based on the expected revenue and relocation costs determined for a vehicle and its corresponding candidate location. It measures the relationship between the expected benefits and execution costs of moving the vehicle from its current location to a candidate location. In some embodiments, the net value of a candidate location may be obtained by subtracting relocation costs from expected revenue. Relocation costs may include energy costs, depreciation costs, and time costs.
[0026] The term "session state" as used in this article refers to the set of data used to maintain recommendation, confirmation, execution, feedback, and reassessment states across events throughout the cruise interaction lifecycle. Session state may include the current interaction phase, vehicle's current location, currently valid recommended location, navigation state, arrival state, set of conditional actions, temporary setting state, waiting timer information, previous round evidence summary, and cooldown timestamp, etc.
[0027] The term "deterministic verification" as used in this article refers to a rigorous review based on preset business rules, precise and repeatable calculation formulas, or deterministic program interfaces. Deterministic verification is used to limit the decision boundaries and output range of large language models, strictly verify the whitelist attribution of recommendation results, the consistency between card and map coordinates, and the authentication legitimacy of setting operations, and block the triggering of subsequent physical navigation or setting adjustment actions when such verification fails.
[0028] Vehicle dispatch recommendations require transforming data from different time and spatial scales into consistent decision inputs. Historical order-taking data reflects the statistical relationship between order-taking locations, travel times, and setting switches under specific spatiotemporal conditions, while the current status reflects vehicle physical location, road network capacity, regional supply and demand, and operational permissions. These two types of data can be recorded using different fields in structured inputs to avoid interpreting offline historical statistics as current real-time facts. When evaluating candidate locations, the energy cost, vehicle depreciation cost, and time cost required for a vehicle to travel to that location can also be considered.
[0029] Therefore, embodiments of this disclosure can logically divide the vehicle dispatch recommendation system into five processing layers: data access, offline experience extraction, online candidate evaluation, constrained decision-making, and execution feedback. The offline experience extraction layer forms structured, searchable experience records from historical order data; the online candidate evaluation layer constructs a candidate whitelist that meets execution constraints based on the current vehicle status and determines the net value of each candidate; the constrained decision-making layer determines recommended locations and associated setting suggestions within the whitelist according to fusion rules; and the execution feedback layer manages arrival condition actions, processes execution receipts, and forms effect samples based on feedback events.
[0030] Figure 1 An exemplary implementation environment 100 for a vehicle dispatch recommendation system is shown. The implementation environment 100 may include a vehicle 110, a terminal device 120, a communication network 130, a server 140, an experience record library 150, and platform data 160.
[0031] Vehicle 110 can be any type of passenger or freight vehicle, such as a ride-hailing vehicle, taxi, or logistics delivery vehicle. Vehicle 110 can be equipped with a positioning device, such as the Global Positioning System (GPS) or the BeiDou Navigation Satellite System, to collect the vehicle's latitude and longitude coordinates, speed, and direction of travel in real time.
[0032] Terminal device 120 can be an in-vehicle unit, in-vehicle central control system, or smart rearview mirror located inside vehicle 110, or it can be a mobile terminal device carried by the driver or operator of vehicle 110, such as a smartphone, tablet, personal digital assistant (PDA), or wearable smart device. Terminal device 120 has a client application installed and running that supports vehicle dispatch recommendation and cruise interaction, and can present a graphical user interface, perform text-to-speech (TTS) broadcasts, and receive touch input, button operation, or voice input from the operator.
[0033] The communication network 130 is used to provide a bidirectional data communication link between the terminal device 120 and the server 140. The communication network 130 can be a wired network or a wireless network, including cellular mobile communication networks, such as fourth-generation mobile communication networks (4G), fifth-generation mobile communication networks (5G) or higher generations, and may also include wireless local area networks (Wi-Fi), Bluetooth connections, satellite communication networks or Internet backbone networks, etc.
[0034] Server 140 can be a single server, a server cluster, a distributed cloud computing platform, or an edge computing node. Server 140 provides unified backend data processing and strategy calculation services for the vehicle dispatch recommendation system. Functionally, server 140 can integrate offline experience processing modules, online candidate evaluation modules, multi-perspective recommendation decision-making modules, execution feedback modules, and deterministic verification modules. It should be understood that these modules are functional units implemented internally by server 140 and do not require separate reference numerals.
[0035] The experience record library 150 communicates with the server 140 to persistently store structured experience records extracted, cleaned, and aggregated from offline historical order data. Each experience record includes clearly defined applicable conditions, order location (including representative coordinates and display name), associated setting status distribution (such as the historical on / off ratio of each receiving switch), and order tag distribution, and supports quick retrieval via multidimensional composite keys during online phases.
[0036] Platform data 160 includes various data sources maintained by the platform or accessed in real time. Logically, platform data 160 is divided into historical data sources and real-time data sources: historical data sources store historical order details, historical service evaluation parameter statistics, and historical order acceptance setting status interval logs; real-time data sources provide current road network topology, real-time road traffic events, regional real-time supply and demand forecast characteristics, route planning distance, estimated time of arrival (ETA), vehicle current battery or fuel status, and current operation permission information. When making recommendation decisions, server 140 can read structured data from experience record library 150 and platform data 160 respectively, and maintain source isolation between historical statistical data and real-time status data in the data pipeline and interface text.
[0037] Figure 2 A flowchart of a vehicle dispatch recommendation interaction process 300 executed by terminal device 120 is shown. Process 300 includes a complete interaction thread encompassing recommendation presentation, setting confirmation, arrival execution, and result presentation. The following is combined with... Figures 3 to 18 The steps and related extended branches of process 300 are described in detail.
[0038] In step 310, the terminal device 120 displays location information and setting suggestions on the cruise interface 200. The location information indicates recommended locations, and the setting suggestions indicate the recommended status of the order acceptance settings, with the recommended status being associated with the recommended locations.
[0039] As a preparatory process before entering the cruise interface, such as Figure 3 As shown, the terminal device 120 can display the cruise function entry interface before entering the cruise function. The cruise function entry card 201 introduces the service feature of the cruise function, which can provide joint recommendations of location and order acceptance settings, to the vehicle operator; the control 202 for entering or returning to cruise is used to receive the trigger operation to start cruise. When the operator clicks the control 202, the terminal device 120 reports a start request to the server 140 and uploads status information such as the vehicle's current physical location and current time period.
[0040] Before generating recommendations or during cruise, such as Figure 4 As shown, operators can also input personalized instructions regarding their order-accepting intentions through the preference input area 240. For example, operators can express preference messages 241 (such as "I want to accept more short orders") through touch input or voice input. The terminal device 120 provides this order-accepting preference instruction to the server 140, so that the server 140 can perform pattern matching between the order-accepting preference instruction and the order-accepting method categories in the experience record, thereby determining the recommended location and appropriate order-accepting settings in subsequent steps.
[0041] After the server 140 completes the recommendation calculation and sends out the recommendation results, the terminal device 120 displays the following: Figure 5 The cruise interface 200 is shown. The cruise interface 200 includes a map display area 210, a location recommendation card 220, a ride-hailing setting suggestion area 230, and a preference input area 240. The map display area 210 indicates the vehicle's current location using a current location marker 211 and indicates the recommended location (e.g., "Location A") using a recommended location marker 212. The location recommendation card 220 displays the recommendation description corresponding to Location A. The ride-hailing setting suggestion area 230 displays ride-hailing setting suggestions associated with Location A.
[0042] In the example shown, the vehicle's current "Special Offer Order Acceptance" switch is off. However, the order tag distribution and historical status distribution corresponding to location A indicate that the conditions for activating the special offer order acceptance are met. Therefore, the setting suggestion indicates that the special offer order acceptance should be activated after arriving at location A. The location recommendation card 220 includes a navigation confirmation control 221 to receive navigation confirmation for destination A; the order acceptance setting suggestion area 230 includes a setting confirmation control 231 to receive a separate confirmation instruction to activate the special offer order acceptance upon arrival.
[0043] In a special recommendation scenario, such as Figure 6 As shown, when the server 140 determines, based on the algorithm evaluation, that the vehicle's current physical location is the recommended location with the best overall score, the terminal device 120 displays a "Stationary Order Acceptance" branch prompt on the cruise interface 200. At this time, the system automatically suppresses the initiation of zero-distance redundant navigation with the current location as the destination coordinate, and no movement path is generated on the map, thereby avoiding sending invalid requests to the underlying navigation engine and eliminating interface route flickering. At the same time, the cruise interface 200 still retains the order acceptance setting suggestion area 230 and the setting confirmation control 231, allowing the vehicle operator to continue to confirm the recommended setting suggestions while stationary, achieving precise setting optimization without the need for displacement.
[0044] In step 320, terminal device 120 receives a confirmation instruction for the setting suggestion. For example, the vehicle operator triggers [something] in the cruise control interface 200. Figure 5 The setup confirmation control 231 is shown. In response to receiving the confirmation instruction, the terminal device 120 uploads the confirmation instruction to the server 140, which stores the confirmation action as a conditional action in the current session state. This conditional action is associated with the location identifier of the recommended location A and the special offer order-taking activation status, and is triggered by the vehicle arriving at the recommended location A.
[0045] After the setting suggestion is confirmed, before the vehicle actually arrives at the recommended location A, such as Figure 7As shown, the order acceptance status display area 232 of the cruise interface 200 displays a state where the setting suggestion has been confirmed but not yet executed, for example, displaying "Confirmed: Activate upon arrival". Since the vehicle has not yet arrived at location A, the current setting status of the special offer order remains unchanged, thereby reducing the possibility of settings taking effect prematurely during driving. The navigation confirmation control 221 can independently receive navigation confirmation for destination A, and the confirmation of setting suggestions and navigation confirmation are processed separately.
[0046] Conversely, if the vehicle operator does not adopt this setting adjustment for the time being, such as Figure 8 As shown, the cruise interface 200 presents a branch state where the setting suggestion has not been adopted. At this time, the operator has not triggered the setting confirmation control 231, and the special order acceptance switch remains in its original off state. Since setting confirmation and navigation confirmation are processed separately, the operator can still independently trigger the navigation confirmation control 221, and the terminal device 120 will normally start the route navigation to the recommended location A, without blocking navigation processing due to the setting suggestion not being confirmed.
[0047] In step 330, based on the confirmation instruction, the terminal device 120, in response to the vehicle arriving at the recommended location, triggers the adjustment of the order acceptance settings to the recommended state.
[0048] When the operator triggers Figure 5 or Figure 7 After the navigation confirmation control 221 in the middle, such as Figure 9 As shown, terminal device 120 enters continuous navigation mode to the recommended location. The cruise interface displays the navigation status display area 250 and the navigation route 251 drawn on the road topology, which precisely connects the current location marker 211 and the recommended location marker 212. During navigation, the order acceptance status display area 232 continuously and clearly indicates the current order acceptance setting status to the operator (preferential order acceptance remains off), ensuring the driver has a clear understanding of their current order acceptance permissions.
[0049] During the journey to the recommended location, the vehicle may deviate from the planned route due to road conditions or driving maneuvers. For example... Figure 10 As shown, when the vehicle's continuous yaw meets the preset correction trigger conditions, the terminal device 120 displays a yaw correction prompt area 252 in the cruise interface and can display correction prompts such as "Deviated from route, replanning". The navigation module regenerates the correction route 251 with the recommended location A as the endpoint, and the recommended location in the session state remains location A. Therefore, route correction does not require frequent switching of the recommended location.
[0050] If, during the vehicle's journey, recommended location A remains within the data validity period, still meets the conditions of road accessibility, traffic restrictions, and preset displacement distance, and the city, date type, time period, service evaluation parameter range, and order acceptance method category used to retrieve the corresponding experience records remain unchanged, and the change in the candidate evaluation results does not reach the re-evaluation threshold, then... Figure 11 As shown, the terminal device 120 maintains the recommended location marker 212 and the original navigation route in the cruise interface, and suppresses repeated broadcasts and full-screen pop-up reminders for location A.
[0051] As the vehicle continues to travel, when its physical location enters a preset geofence based on the navigation coordinates of the recommended location A, the terminal device 120 determines that an arrival event has been established and sends an arrival event to the server 140. The server 140 retrieves the confirmed conditional actions associated with location A from the session state, verifies the execution status and operation permissions of the actions, and then initiates a switch adjustment request to the underlying deterministic setting interface.
[0052] like Figure 12 As shown, before receiving a successful execution receipt from the settings interface, the order status display area 232 displays a processing status such as "Arrived" or "Special Offer Order Adjustment in Progress," without prior notification of successful adjustment. For the stationary order acceptance branch where the recommended location is the vehicle's current location, upon receiving a confirmation instruction for the settings suggestion, the arrival condition can be directly determined as met, and the adjustment can be initiated according to the same action status verification and permission check process.
[0053] In step 340, in response to receiving an execution receipt indicating that the order acceptance settings have been adjusted to the recommended state, the terminal device 120 displays a successful adjustment message on the navigation interface. Figure 13 As shown, after the underlying system completes the switch status update and provides the execution receipt to the terminal device 120, the setting adjustment success prompt area 233 displays "Setting adjustment successful". Therefore, the adjustment success status displayed on the interface is based on the execution receipt of the underlying setting interface, and is distinguished from the intention status of the setting suggestion being confirmed and the status of adjustment processing.
[0054] After step 340 is completed, the vehicle will cruise and wait for orders normally at the recommended location A under the optimized settings (special offer enabled). Process 300 further covers various subsequent business expansion interaction branches: Firstly, combining Figure 14When a vehicle arrives at recommended location A and has not received any orders for an extended period, reaching a first time threshold (the first time threshold is a configurable engineering parameter, example being approximately 5 minutes), the terminal device 120 displays a waiting prompt area 260 on the cruise interface. The waiting prompt is determined based on at least one of the following: real-time queuing capacity information around the current location obtained by the server 140, regional supply and demand forecast gap, or historical average waiting time in the business district. For example, it may display a prompt such as "5 minutes have passed, current regional demand is stable, it is recommended to continue waiting for orders in place," to synchronize the current waiting status and expected waiting information with the vehicle operator.
[0055] Secondly, in combination Figure 15 If the waiting time of a vehicle at location A reaches the second time threshold, which is greater than the first time threshold, or if the data validity period of location A expires, location A no longer meets the executability verification, or the decline in the candidate evaluation result of location A relative to the previous evaluation result reaches the re-recommendation threshold, the server 140 will start a new round of candidate evaluation and provide the updated recommendation result to the terminal device 120.
[0056] like Figure 15 As shown, the updated recommendation prompt area 261 displays a reassessment explanation, the recommendation marker 212 in the map display area is updated to the new location B, and new setting suggestions corresponding to location B are presented. Navigation and setting suggestions for location B require a new confirmation instruction and do not reuse the previous confirmation instruction for location A; previously unexecuted conditional actions are marked as invalid to maintain consistency between the old and new recommendation states.
[0057] Thirdly, in combination Figure 16 When a vehicle successfully matches a real order at the recommended location, the terminal device 120 exits the cruise recommendation interface and switches to the order acceptance and pick-up status interface. The interface displays an order status prompt area 270, a pick-up location marker 271 guiding the vehicle to the actual passenger pick-up point, a pick-up route 272, and an order information area 273 displaying passenger needs and business attributes. It should be clear that the pick-up location marker 271 indicates the passenger pick-up point for a specific order, and has a completely different business meaning and data source from the location marker 212 recommended by the system algorithm during the cruise phase. The two are strictly distinguished within the system by different fields.
[0058] Fourth, in combination Figure 17 Once the trip is completed and the order is in the "completed" state, the terminal device 120 displays a trip end and completion feedback interface. The order information area 273 can display completion results such as "Order Completed" and "Revenue Recorded." The interface can also provide controls 202 to enter or return to the cruise, allowing vehicle operators to start the next cruise after completion.
[0059] Fifth, in combination Figure 18 As an independent optional branch, if this scheduling task is associated with scheduling rights set by the platform, and the server 140 receives a settlement result indicating that the corresponding rights conditions have been met, the terminal device 120 can display the rights settlement feedback area 280 in the cruise interface. Rights settlement feedback is a notification of the rights settlement result, and its triggering basis is... Figure 13 The execution receipt indicates that the order settings have been successfully adjusted, which is different from the one in the middle.
[0060] Figure 19 A flowchart of the vehicle dispatch recommendation process 400 executed by server 140 is shown. Process 400 encompasses the complete technical implementation of offline experience distillation, online constrained evaluation, multi-perspective fusion decision-making, and closed-loop feedback. The following is combined with... Figure 19 and Figure 20 The steps of process 400 are explained in detail.
[0061] In step 410, the server 140 obtains experience records and vehicle status information. The experience records are obtained based on historical order data statistics and indicate the applicable conditions, as well as the order location and order setting status corresponding to the applicable conditions.
[0062] To construct an experience record capable of being retrieved and calculated according to structured conditions, the offline data processing module can perform the following processing: First, the system prepares reference samples and determines order-taking methods. Each city is treated as an independent processing unit, and different historical datasets are formed based on the preset numerical ranges of service evaluation parameters. Within each historical dataset, the system determines the completeness of records for each statistical object's unit online time revenue, number of valid orders, location, and order-taking settings within a preset historical statistical window. Historical order-taking records with unit online time revenue not lower than a preset revenue percentile, a number of valid orders reaching a quantity threshold, and a record completeness threshold are selected as reference samples for forming experience records.
[0063] In one example, the income percentile could be the 80th percentile, thus selecting a reference sample from the top 20% of the corresponding income distribution. The system can categorize the reference sample into a first order-taking category and a second order-taking category based on historical order-taking behavior statistics, such as the number of orders received per unit time and the distribution of order distances. The first order-taking category corresponds to historical order-taking behavior with higher frequency and shorter average order distances, while the second order-taking category corresponds to historical order-taking behavior with longer average order distances.
[0064] In another embodiment, the system can generate an order-accepting behavior feature vector based on historical order origin, order setting status, order tags, waiting time, empty driving distance, and unit order revenue. The system then performs scale transformation on each feature dimension and uses unsupervised clustering to form multiple order-accepting method categories. Preset rule classification and unsupervised clustering are different embodiments and do not need to be executed simultaneously.
[0065] Second, order parsing is aligned with switch / off time periods. For each order in the historical order data and the asynchronously recorded business switch / off status logs, the system extracts the order acceptance time and starting coordinates for each order. Following the rule that "the start time of the status interval is no later than the order acceptance time, and the end time of the status interval is later than the order acceptance time," the system matches the order with the corresponding valid switch / off status intervals. The system records different business switches separately. For statistical processing, an on state is recorded as 1; a off state is recorded as 0. For missing or invalid values, a preset conservative statistical rule can be used to record them as 0, but this statistical code does not indicate that the actual state of the underlying switch is confirmed as off. After matching, each order forms an order acceptance record including the order acceptance time, starting coordinates, order tag, and statistical status code.
[0066] Third, source location grid quantization and display name cleaning. To prevent existing Point of Interest (POI) identifiers on the platform from becoming conflated across regions, leading to coordinate and name errors, the system spatially quantizes the latitude and longitude coordinates of the starting point of each order record by retaining a preset number of bits, thus mapping them to fixed-precision spatial grid units. The city identifier is then concatenated with these grid coordinates to construct a unique source location identifier within the city. For order records belonging to the same source location identifier, the representative coordinates are determined by weighting the number of orders.
[0067] To address noise in the original order address text, the address cleaning module filters out invalid or vague names such as broad administrative divisions. It determines the name specificity score based on the frequency of the name in historical orders, the text length, and whether it contains specific location features. Names with high frequency and name specificity reaching the threshold are selected as display names. When a reliable name cannot be obtained, empty names or unknown markers are retained to avoid using internal grid codes as display names.
[0068] Fourth, the two-stage spatial aggregation generates order-receiving locations. In the first stage, the system counts the order volume of each source location, identifies source locations with order volumes greater than the seed quantity threshold as core locations, and sequentially uses unassigned core locations as aggregation centers in descending order of order volume. Other unassigned core locations whose distance from the aggregation center is no greater than the first distance threshold are then assigned to the aggregation center to form order-receiving locations.
[0069] During the selection of aggregation centers, the system determines the coverage ratio based on the cumulative order volume of core locations already used to form order-taking locations and the total number of valid orders in the corresponding city. It can stop selecting new aggregation centers when the coverage ratio reaches a coverage threshold. In the second stage, for source locations throughout the city that have not yet been assigned, including core locations that have not yet served as aggregation centers due to reaching the coverage threshold and other source locations, the system determines the order-taking location closest to that source location. If the distance is not greater than a second distance threshold, the source location is assigned to the nearest order-taking location; otherwise, it is excluded from hotspot experience. The second distance threshold is greater than the first distance threshold.
[0070] The aggregated representative coordinates of each order receiving location can be determined based on the aggregation center and the order volume of the source location, and then determined according to the weight and the corresponding coordinates; the display name of the order receiving location can be selected from the display name of the source location with the highest order volume and a valid name.
[0071] Fifth, multi-dimensional sharding and composite key index construction. The system shards the spatially aggregated order records according to city identifier, service evaluation parameters, date type, time period, and order method category. The system constructs composite keys based on the above dimensions and stores each shard in association with the composite key. For preset combinations where no valid order locations are found, the system can retain empty shards indexed by the corresponding composite key to distinguish between the state of not finding hotspots and the state of abnormal retrieval.
[0072] Sixth, the system calculates the distribution of order acceptance settings and order tag distribution. The system determines the proportion of orders with specific settings in a given state based on order criteria. Specifically, the system counts the number of valid orders with a specific order acceptance setting enabled in the same segment as the first order quantity, and counts the total number of valid orders in that segment used to count that order acceptance setting as the second order quantity. The enabling ratio is determined based on the ratio of the first order quantity to the second order quantity.
[0073] For missing or invalid order acceptance status records, missing records can be handled according to preset statistical rules; the corresponding statistical processing results do not indicate that the actual order acceptance status corresponding to the missing record has been determined to be closed. For each order acceptance location within a segment, the system determines the order tag distribution based on the number of orders matching the corresponding order tag and the total number of orders for that order acceptance location. The order acceptance status distribution belongs to the segment-level statistical results, while the order tag distribution belongs to the order acceptance location-level statistical results.
[0074] Seventh, waiting parameters and periodic updates. The system extracts waiting times from historical order data to determine the distribution or reference range of waiting times for different order types under different spatiotemporal conditions, which serves as the waiting time condition in the applicable conditions. The experience record library 150 can use newly added historical data according to a preset update cycle to re-execute the matching of order time and status interval, the formation of order records, spatial aggregation, and segmented statistics to update the experience records.
[0075] In an embodiment where the waiting parameter is used as an applicable condition, the server 140 obtains the historical order waiting records corresponding to the current vehicle and determines the vehicle's waiting parameters based on the waiting time statistics in the historical order waiting records. For example, the waiting parameters can be determined based on the average waiting time, median waiting time, or waiting time distribution within a preset historical time range. The server 140 matches the waiting parameters with the waiting time conditions corresponding to the order acceptance method category in the experience records and uses the matching results to filter or sort the experience records. When the vehicle's historical order waiting records do not meet the preset data volume condition, the waiting parameter can be set to missing and is not used as a mandatory matching condition.
[0076] In step 420, the server 140 determines candidate locations based at least on the matching results of the state information and the applicable conditions. The candidate locations satisfy the execution constraints for vehicles.
[0077] When vehicle 110 is in cruise mode, server 140 receives vehicle status information uploaded by terminal device 120, including the vehicle's current latitude and longitude coordinates, current city, current time period, current date type, the numerical range of service evaluation parameters, vehicle type and energy cost parameters, current order setting status, and related operation permissions.
[0078] Server 140 will include the city identifier, service evaluation parameter range, date type, time period, and other information from the status information. Figure 4 The order acceptance method category combination matched by the preference message in the system serves as the online search key, retrieving the corresponding experience record from the experience record database 150. Simultaneously, the server 140 performs a spatial range search in the platform data 160 based on the vehicle's current location to obtain the real-time supply and demand location.
[0079] The server 140 converts the order-receiving locations and real-time supply and demand locations in the experience records into a standardized candidate structure. The candidate structure includes candidate identifier, location name, representative coordinates, spatial straight-line distance, estimated travel time, source evidence type, and missing item flags.
[0080] After generating unified candidates, server-side 140 performs integrity and executability checks on the locations to be filtered. Integrity checks verify the completeness of the location name and coordinate fields. Executability checks may include at least one of the following: road accessibility, traffic restrictions, displacement distance limits, data validity period, and operation permissions.
[0081] Server 140 determines navigation coordinates for locations that pass the executability check and associates the display name, aggregated representative coordinates, and navigation coordinates using the same location identifier. It recommends using the appropriate display name, map markers, and navigation coordinates, and that arrival events be determined based on the positional relationship between the vehicle's location and the navigation coordinates. Locations that pass both integrity and executability checks are identified as candidate locations and form a candidate whitelist.
[0082] For locations to be screened based on regional supply and demand data but lacking corresponding experience records, the system sets the experience data to missing and the experience matching score to zero to avoid presenting non-existent historical experience as valid historical statistical evidence. Because of the lack of corresponding historical setting status distribution for determining the recommended order-accepting settings, the system does not generate setting adjustment suggestions or corresponding conditional actions for this location to be screened. However, it can provide location information or navigation suggestions if the location meets other recommended conditions.
[0083] In step 430, server 140 determines the recommended location and order-taking status based on the matching results and the net value of candidate locations. The recommended location is selected from candidate locations, the recommended status is determined based on the recommended location and the order-taking status indicated by experience records, and the net value is determined based on the expected revenue and displacement costs for the vehicle and the corresponding candidate location.
[0084] For each candidate location i in the candidate whitelist, server 140 combines the current physical location of the vehicle with the real road network topology between the candidate location to obtain the path planning distance L. i ETA (Estimated Travel Time) i Displacement cost C i It consists of three components: energy cost, depreciation cost, and time opportunity cost: Server-side 140 is based on path distance L i Energy costs are calculated by combining the vehicle's energy consumption type (fuel consumption per 100 kilometers for gasoline vehicles or electricity consumption per 100 kilometers for electric vehicles) and the unit price of energy; based on the path distance L. i Calculate vehicle wear and tear depreciation costs based on vehicle model and mileage depreciation rate; ETA based on estimated driving time. i The opportunity cost of time is calculated by combining the driver's revenue per unit of online time (hourly wage level) during similar historical periods. The energy cost, depreciation cost, and time cost are then added together to obtain the comprehensive displacement cost C required to reach candidate location i. i .
[0085] Expected revenue R of candidate locations i This can be determined based on regional supply and demand data and order revenue statistics under similar or comparable spatiotemporal conditions within a historical time window. Server-side 140 determines the expected revenue R for each candidate location. i and displacement cost C i And can reduce expected income R i Subtract displacement cost C i The net value V of the candidate site is obtained. i When there is insufficient data to determine expected revenue, the corresponding revenue item is marked as missing, and subsequent processing is performed according to the preset missing data handling rules.
[0086] To compare multiple candidate locations in the candidate whitelist, server 140 determines multiple scoring items according to preset scaling rules. In one embodiment, the scoring items include an expert experience matching score. i Driver Value Score D i Expert Methodology Division M i and general strategy G i The fusion evaluation result F of candidate location i i You can follow F i =0.25E i +0.50D i +0.15M i +0.10G i Confirmed. Expert experience matching score E i Characterizes the degree of matching between historical statistical information of candidate locations and vehicle status information and applicable conditions; driver value score D i Determined at least based on the net value of candidate sites; expert methodology M i and general strategy G i These respectively characterize the degree of matching between candidate locations and preset order acceptance methods and general recommendation conditions.
[0087] If any data required for a scoring item is missing, the system sets a corresponding missing flag, sets the scoring item to zero, and keeps the weights of each scoring item unchanged, without transferring the weights of the missing items to other scoring items.
[0088] The server-side 140 can set preset revenue conditions, such as requiring the net value of candidate locations to be greater than zero or to reach a preset lower limit of net value, and then select candidate locations from the candidate whitelist that meet the preset revenue conditions to form a recommended candidate set. Only candidate locations in the recommended candidate set can participate in subsequent fusion comparisons.
[0089] The server-side component 140 then generates structured input for the Large Language Model (LLM). This structured input includes a candidate whitelist, location identifiers for candidate locations, display names, coordinates, source type, net value, various scoring components, fusion evaluation results, historical statistical evidence fields, current state evidence fields, missing data flags, fusion rules, and output constraints. The output constraints restrict the LLM to selecting candidate locations only from the whitelist and outputting the location identifiers of the selected locations and a recommendation description generated based on the structured evidence.
[0090] The large language model does not change the scoring items, item weights, candidate whitelist, or set operation permissions. When the fusion evaluation results of multiple candidate locations are the same, the deterministic processing program determines one of the candidate locations according to pre-set and repeatable parallel processing rules. The large language model does not autonomously determine the priority order of parallel candidates based on the order of candidate locations in the input. When the candidate whitelist is empty, the server 140 does not generate recommended locations or set adjustment suggestions, but provides a status description based on vehicle status information.
[0091] After the large language model generates recommendation results, the deterministic verification module recalculates the fusion evaluation results of the corresponding candidate locations and verifies whether the selected candidate locations belong to the candidate whitelist, whether the recommendation description and location information correspond to the same location identifier and coordinates, whether the setting suggestions are generated by preset setting rules, and whether the corresponding setting operations meet the operation permissions. If the deterministic verification passes, the server 140 provides the corresponding recommendation results; if the deterministic verification fails, the server 140 blocks the navigation operation and order setting adjustment operation corresponding to the recommendation result and provides a status description.
[0092] Only when the deterministic verification passes will the server 140 execute step 440 and provide the corresponding recommendation result to the terminal device 120.
[0093] If the deterministic validation fails, the server will block navigation operations and order setting adjustment operations associated with the recommendation result (version 140). Figure 20 As shown, the cruise interface can display a status description area 290 indicating that suggestions cannot be executed. The server 140 generates a status description based on the vehicle's current status and does not provide corresponding navigation instructions or setting adjustment instructions to reduce the impact of abnormal recommendation results on vehicle driving and order acceptance settings.
[0094] Regarding the determination of the recommended status for setting suggestions, the recommendation determination module reads the order tag distribution corresponding to the recommended location, the historical order acceptance setting status distribution corresponding to the spatiotemporal segment where the recommended location is located, the current order acceptance setting status of the vehicle, and the current operation permissions. When the proportion of a specific order tag for the recommended location reaches a tag proportion threshold, and the proportion of the specific order acceptance setting enabled corresponding to the spatiotemporal segment reaches a preset enabling threshold, if the vehicle's order acceptance setting is currently disabled and the corresponding operation permissions allow it to be enabled, the server 140 will enable the order acceptance setting and determine it as the recommended status, and generate corresponding setting suggestions based on the location information.
[0095] In step 440, the server 140 provides the terminal device with location information and setting suggestions. The location information indicates the recommended location, and the setting suggestions indicate the recommended status.
[0096] After providing the recommendation results to the terminal device 120, the server 140 manages the session state corresponding to this round of recommendations. Upon receiving a confirmation instruction for the setting suggestion, the server 140 writes the corresponding conditional action into the session state. The conditional action is associated with the recommended location and the target setting state, with the arrival of the vehicle at the recommended location as the trigger condition, and the execution content being to adjust the order acceptance settings to the target setting state.
[0097] When an arrival event indicating that a vehicle has reached a recommended location is received, server 140 retrieves the corresponding conditional action from the session state, verifies the recommended location, target setting status, and operation permissions, and calls the deterministic setting interface upon successful verification. After receiving an execution receipt from the setting interface indicating that the order acceptance settings have been adjusted to the target setting status, server 140 updates the session state and provides adjustment success information to terminal device 120. Before receiving the corresponding execution receipt, the conditional action is not recorded as successfully executed.
[0098] The server 140 can also maintain the recommendation status based on the cruise event information provided by the terminal device 120. When the recommended location is still within the data validity period, still meets the conditions of road accessibility, traffic restrictions, and preset displacement distance, and the city, date type, time period, service evaluation parameter range, and order acceptance method category used for retrieving experience records have not changed, and the change in the candidate evaluation results has not reached the re-evaluation threshold, the server 140 maintains the current recommended location and prevents the repeated provision of recommended broadcast content for that recommended location to the terminal device 120. When the continuous deviation of the vehicle meets the correction trigger condition, the server 140 provides correction information to the terminal device 120 so that the terminal device 120 can generate or update the route with the recommended location as the destination, and maintain the current recommended location.
[0099] During the vehicle dispatch interaction, server 140 can receive setting execution receipts, arrival events, order acceptance events, location records, order completion results, and order data corresponding to the recommendation results. Setting execution receipts are used to update the execution status of conditional actions and order acceptance settings, and are not used to separately determine waiting time, empty driving distance, or order revenue. When both arrival and order acceptance events exist and have corresponding event times, server 140 determines the waiting time based on the event times of both. When a location record exists corresponding to the time interval during which no order was received, server 140 determines the empty driving distance based on the location record. When an order completion result and corresponding order data exist, server 140 determines the order revenue based on the order data.
[0100] The server 140 records at least one of the following—waiting time, empty driving distance, and order revenue—determined based on the corresponding feedback data, as an effect sample. For recommendation results that use experience records to determine recommended locations, the server 140 associates and stores the effect sample with the corresponding experience record and updates the effect statistics corresponding to that experience record. If data for determining the corresponding effect metric is missing, that effect metric is not generated.
[0101] The following example of candidate location comparison illustrates the relationship between historical statistical evidence, real-time status evidence, displacement costs, and executability constraints.
[0102] In one specific embodiment, after the vehicle enters cruise mode, the server 140 recalls candidate location A from the experience record based on the vehicle's current location and current applicable conditions. Simultaneously, based on statistical data and real-time supply and demand data near the current location, candidate locations B and C are obtained, and the three locations are converted into a unified candidate structure. Candidate location A has strong historical evidence of high-frequency short-distance orders, but its displacement distance from the vehicle's current location is large; candidate location B has a lower historical opportunity index than candidate location A, but its displacement cost is lower; candidate location C has real-time supply and demand data, but fails the road accessibility verification. The server 140 excludes candidate location C from the candidate whitelist and determines the expected revenue, displacement cost, net value, and experience matching results for candidate locations A and B respectively.
[0103] When the net value gain of candidate location A can cover the increased displacement cost relative to candidate location B, and the fusion evaluation result of candidate location A meets the preset selection criteria, the server 140 can determine candidate location A as the recommended location; otherwise, candidate location B will be determined as the recommended location, and a recommendation description of waiting for orders at the current location or nearby can be provided. The recommendation description, location information, and navigation operations use the same display name and coordinates associated with the same location identifier. Corresponding setting suggestions are only generated when the order tag distribution, historical order acceptance setting status distribution, current order acceptance setting status, and operation permissions corresponding to the recommended location all meet the preset setting rules.
[0104] Based on the same inventive concept as the above method, embodiments of this disclosure also provide corresponding apparatus for implementing the above method.
[0105] In some embodiments, an interactive device for vehicle dispatch recommendation is provided, disposed in a terminal device 120. The interactive device includes a first presentation module, a receiving module, a triggering module, and a second presentation module. The first presentation module is configured to present location information and setting suggestions in the cruise interface. The location information indicates a recommended location, and the setting suggestions indicate a recommended status for order acceptance settings, with the recommended status associated with the recommended location. The receiving module is configured to receive a confirmation instruction for the setting suggestions. The triggering module is configured to, based on the confirmation instruction, trigger an adjustment of the order acceptance settings to the recommended status in response to the vehicle arriving at the recommended location. The second presentation module is configured to, in response to receiving an execution receipt, present an adjustment success prompt in the cruise interface, with the execution receipt indicating that the order acceptance settings have been adjusted to the recommended status.
[0106] In some optional embodiments, the setting suggestion is determined based on the order tag distribution corresponding to the recommended location, the historical order acceptance setting status distribution corresponding to the spatiotemporal segment where the recommended location is located, the current status of the order acceptance setting, and the operation permissions. The recommended status may include an enabled status, which corresponds to a specific receiving switch in the order acceptance settings; when the historical enabled ratio of the receiving switch in the corresponding spatiotemporal segment reaches the enabled threshold, the order tag distribution, the current status, and the operation permissions can be combined to determine whether to generate an enabled suggestion.
[0107] In some alternative embodiments, the first presentation module and the triggering module are further configured to: present map markers and recommended descriptions in the cruise interface, the map markers and recommended descriptions corresponding to recommended locations; and trigger navigation with the coordinates corresponding to the map markers as the destination coordinates in response to receiving a navigation confirmation instruction.
[0108] In some optional embodiments, the first presentation module is further configured to: in response to the vehicle's continuous yaw meeting the correction trigger condition, present a correction prompt in the cruise interface, guide the vehicle to travel along a route with the recommended location as the destination, and maintain the recommended location corresponding to the map marker; in response to the recommended location still being within the data validity period, still meeting the executability verification, and the change in the candidate evaluation result not reaching the re-evaluation threshold, maintain the location information in the cruise interface and suppress repeated recommendation broadcasts for the recommended location; in response to the vehicle arriving at the recommended location and the waiting time without receiving an order reaching a first time threshold, present a waiting prompt in the cruise interface; in response to the waiting time reaching a second time threshold, or the recommended location no longer meeting the executability verification, present updated location information and setting suggestions in the cruise interface; and in response to the recommended location being the vehicle's current location, present a "wait for orders in place" prompt in the cruise interface and suppress navigation with the current location as the destination.
[0109] In some optional embodiments, the receiving module is further configured to: receive an order acceptance preference indication; and send the order acceptance preference indication to the server so that the server can determine the recommended location and recommendation status based on the matching result of the order acceptance preference indication and experience records.
[0110] In other embodiments, a vehicle dispatch recommendation device set on a server is provided. The device includes an acquisition module, a candidate determination module, a recommendation determination module, and a provision module. The acquisition module is configured to acquire experience records and vehicle status information. The experience records are statistically derived from historical order-taking data and indicate applicable conditions, as well as the order-taking locations and order-taking settings corresponding to the applicable conditions. The candidate determination module is configured to determine candidate locations based at least on the matching results of the status information and the applicable conditions. The candidate locations satisfy the execution constraints for the vehicle. The recommendation determination module is configured to determine recommended locations and recommended order-taking settings based on the matching results and the net value of the candidate locations. The recommended locations are selected from the candidate locations, and the recommended status is determined based on the recommended locations and the order-taking setting status indicated by the experience records. The net value is determined based on the expected revenue and displacement costs determined for the vehicle and the corresponding candidate locations. The provision module is configured to provide location information and setting suggestions to terminal devices.
[0111] In some optional embodiments, the acquisition module is configured to acquire experience records by: for different orders in historical order data, matching the status interval of the order setting based on the corresponding order acceptance time, wherein the start time of the status interval is no later than the order acceptance time and the end time of the status interval is later than the order acceptance time; associating the order acceptance time, starting position, order tag, and the order setting status indicated by the matched status interval for each order to obtain an order acceptance record; performing aggregation processing on the order records to obtain statistical results of the order acceptance location and the status distribution indicating the order setting; and associating and storing the order acceptance location, statistical results, and applicable conditions to obtain an experience record.
[0112] In some optional embodiments, the applicable conditions include spatiotemporal conditions, and the status information includes the vehicle's current location and current time period. The acquisition module performs aggregation processing on order records, including: segmenting the order records based on date type, time period, service evaluation parameters, and order acceptance method category, and statistically analyzing the status distribution of order acceptance locations and settings for each segment; constructing a composite key based on the city identifier and the date type, time period, service evaluation parameters, and order acceptance method category, and storing the experience records corresponding to the segment in association with the composite key; when recommending vehicle dispatch, constructing a search key based on the city identifier corresponding to the current location and the corresponding date type, time period, service evaluation parameters, and order acceptance method category, and obtaining experience records matching the search key. For preset combinations that do not contain order acceptance locations, empty segments indexed by the composite key are retained.
[0113] In some optional embodiments, the acquisition module is further configured to perform two-stage spatial aggregation in the following manner: quantizing the latitude and longitude of the starting point location with a predetermined precision to obtain grid coordinates; constructing source location identifiers based on grid coordinates and city identifiers; determining source locations and corresponding order quantities based on source location identifiers, and identifying source locations with order quantities exceeding a seed quantity threshold as core locations; in descending order of order quantity, using unassigned core locations as aggregation centers, incorporating other unassigned core locations whose distance from the aggregation center is no greater than a first distance threshold into the aggregation center to form order-receiving locations, and marking the aggregation center and the incorporated core locations as assigned; Based on the cumulative number of orders corresponding to the core locations already used to form order-receiving locations and the total number of valid orders in the corresponding cities, a coverage ratio is determined, and in response to the coverage ratio reaching a coverage threshold, the selection of new aggregation centers is stopped; for source locations that have not yet been assigned, the order-receiving location closest to the source location is determined, and if the distance is not greater than a second distance threshold, the source location is assigned to the nearest order-receiving location, and if the distance is greater than the second distance threshold, the source location is excluded from the order-receiving locations used to form hotspot experience, where the second distance threshold is greater than the first distance threshold; and the coordinate fusion weight is determined based on the number of orders of the aggregation center and the source location respectively, and the coordinates of the order-receiving location are determined based on the weight and the corresponding coordinates.
[0114] In some optional embodiments, the candidate determination module is configured to: obtain locations to be screened, including order-taking locations selected based on matching results and locations determined based on regional supply and demand data; convert locations to be screened from different sources into candidate information in a unified format; verify the integrity of the name and coordinates based on the candidate information, and perform executability verification on the locations to be screened according to road accessibility, traffic restrictions, displacement distance restrictions and data validity period; and determine the locations to be screened that pass the integrity verification and executability verification as candidate locations, and form a candidate whitelist.
[0115] In some optional embodiments, the recommendation determination module is configured to: obtain the path distance and estimated travel time between the candidate location and the vehicle's current location; determine the energy cost and depreciation cost based on the path distance and the vehicle's energy cost and depreciation cost parameters; determine the time cost based on the estimated travel time and the historical unit-time revenue associated with the vehicle; determine the displacement cost based on the energy cost, depreciation cost, and time cost; determine the expected revenue based on regional supply and demand data and order revenue statistics under the same or similar spatiotemporal conditions within a historical time window; and determine the net value based on the expected revenue and displacement cost.
[0116] In some optional embodiments, the recommendation determination module is configured to: generate structured input, which includes a candidate whitelist, net value of candidate locations, expert experience matching score, driver value score, expert methodology score, general strategy score, fusion rules, historical statistical evidence field, current state evidence field, missing data flag, and output constraints; process the structured input using a large language model to obtain recommendation results; and determine recommended locations and recommendation descriptions based on the recommendation results. When data for any scoring item is missing, a corresponding missing data flag is set, the corresponding scoring item is set to zero, and the weights of each scoring item remain unchanged.
[0117] In some optional embodiments, the recommendation determination module is further configured to perform deterministic verification on the recommendation results, including: re-determining the fusion evaluation result corresponding to the recommendation result according to the fusion rules; verifying whether the candidate location indicated by the recommendation result belongs to the candidate whitelist; verifying whether the recommendation description, location information and recommendation result correspond to the same location identifier, name and coordinates; and verifying whether the setting suggestions are generated by preset setting rules and whether the corresponding adjustment operations meet the operation permissions. In response to the deterministic verification passing, the corresponding location information and setting suggestions are provided to the terminal device; in response to the deterministic verification failing, navigation operations and order acceptance setting adjustment operations associated with the recommendation result are blocked, and a status description generated based on vehicle status information is provided.
[0118] In some optional embodiments, the providing module and the recommendation determining module are further configured to: receive a setting suggestion confirmation instruction from the terminal device; store conditional actions associated with the recommended location and recommended status in the session state, wherein the conditional actions are triggered by the vehicle arriving at the recommended location and the execution content is to adjust the order acceptance setting to the recommended status; in response to receiving an arrival event, retrieve the corresponding conditional action from the session state and trigger its execution; in response to receiving an execution receipt indicating that the order acceptance setting has been adjusted to the recommended status, provide the terminal device with adjustment success information; receive the arrival event, order acceptance event, location record, order completion result, and order data corresponding to the recommendation result; determine the waiting time based on the event time of the arrival event and the order acceptance event; determine the empty driving distance based on the location record corresponding to the time interval of no order received; determine the order revenue based on the order completion result and the order data; and record at least one of the waiting time, empty driving distance, and order revenue determined based on the corresponding feedback data as an effect sample associated with the experience record.
[0119] The modules included in the aforementioned interactive and recommendation devices can be implemented using software, hardware, firmware, or any combination thereof. In some embodiments, one or more modules can be implemented using machine-executable instructions stored on a storage medium. In addition to or as an alternative to machine-executable instructions, some or all of the units in the device can be implemented using field-programmable gate arrays, application-specific integrated circuits, system-on-a-chip, or other hardware logic components.
[0120] Figure 21 A block diagram of an exemplary electronic device 1100 capable of implementing embodiments of the present disclosure is shown. The electronic device 1100 may be used to implement the terminal device 120 or server 140 discussed above.
[0121] like Figure 21 As shown, the electronic device 1100 includes one or more processing units or processors 1110, a memory 1120, a storage device 1130, one or more communication units 1140, one or more input devices 1150, and one or more output devices 1160. The processor 1110 can be an actual physical processor or a virtual processor, and is capable of executing various data processing and strategy logic according to program instructions stored in the memory 1120. In a multiprocessor system, multiple processors execute instructions in parallel to improve the parallel processing throughput of the electronic device 1100.
[0122] Memory 1120 may be volatile memory, such as registers, caches, or random access memory, or non-volatile memory, such as read-only memory, electrically erasable programmable read-only memory, or flash memory, or a combination of the above. Storage device 1130 may be a removable or non-removable computer-readable storage medium, such as a disk, solid-state drive, optical disc drive, or cloud storage array, capable of storing program code, map data, experience records, and model parameters in a non-transient form. Memory 1120 may contain a computer program product 1125, which includes one or more program modules configured to perform the method procedures described herein.
[0123] Communication unit 1140 enables bidirectional network communication with other external electronic devices via a communication medium. Communication unit 1140 can support interfaces such as cellular wireless networks, Ethernet, Wi-Fi, or Bluetooth. Electronic device 1100 can access the cloud environment or vehicle network system via communication unit 1140 in the form of a single computing node, computing cluster, or distributed network. Input device 1150 may include a touchscreen, physical buttons, microphone, or positioning antenna; output device 1160 may include a display screen, head-up display, speaker audio unit, or vibration motor, used to provide visual, auditory, or tactile interactive feedback.
[0124] Embodiments of this disclosure also provide a computer-readable storage medium having computer-executable program instructions stored thereon. When executed by a processor, the computer-executable instructions cause the processor to implement the methods described above for a terminal device or server. The computer-readable storage medium may be in the form of a non-transitory storage medium, disposed within an electronic device or communicatively connected to an electronic device via a network.
[0125] Embodiments of this disclosure also provide a computer program product including computer-executable instructions. When the computer-executable instructions are run on or executed by a processor, they cause the processor to implement the methods or processes described above. The computer program product can be implemented as a client application running on a mobile terminal, a software development kit (SDK) integrated into an in-vehicle infotainment system, a distributed recommendation engine service running on a server, a microservice plugin, or a cloud scheduling package, etc.
[0126] The flowcharts and block diagrams in the accompanying drawings illustrate possible implementation architectures, functions, and operations of systems, methods, and computer program products according to embodiments of this disclosure. The blocks in the flowcharts or block diagrams may represent a portion of a module, program segment, or instruction. In the absence of causal dependencies, some processes may be executed in parallel or in a different order. For embodiments employing the mainline of "confirm setting suggestion—vehicle arrives at recommended location—trigger setting adjustment—receive execution receipt and present adjustment success," the aforementioned events are processed according to the stated causal relationship to maintain the correspondence between session state and underlying execution results. The above timing description is for embodiments employing this interaction mainline and does not require other embodiments to include all events.
[0127] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed implementations. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described implementations. The terminology used in this disclosure is chosen to best explain the principles, practical applications, or improvements to the prior art of the implementations, or to enable others skilled in the art to fully understand the various implementations disclosed herein.
Claims
1. An interactive method for vehicle dispatch recommendation, executed by a terminal device, comprising: The cruise interface displays location information and setting suggestions. The location information indicates recommended locations, and the setting suggestions indicate the recommended status of the order acceptance settings. The recommended status is associated with the recommended locations. Receive confirmation of the suggested settings; Based on the confirmation instruction, in response to the vehicle arriving at the recommended location, the order acceptance settings are adjusted to the recommended state. as well as Upon receiving the execution receipt, a success message is displayed on the navigation interface, indicating that the order acceptance settings have been adjusted to the recommended state.
2. The method according to claim 1, wherein the setting suggestion is determined based on the order tag distribution corresponding to the recommended location, the status distribution of historical order acceptance settings, the current status of the order acceptance settings, and the operation permissions, and the status distribution is associated with the spatiotemporal conditions corresponding to the recommended location.
3. The method according to claim 2, wherein the recommended state includes an on state, the on state corresponds to a receiving switch in the order receiving settings, the receiving switch is used to control the receiving of at least one order type, and the state distribution indicates that the on state of the receiving switch has reached an on threshold.
4. The method according to claim 1, further comprising: The cruise interface displays map markers and recommended descriptions, which correspond to the recommended locations indicated by the location information. as well as In response to receiving a navigation confirmation instruction, navigation is triggered with the coordinates corresponding to the map marker as the destination coordinates.
5. The method according to claim 4, further comprising: In response to the continuous yaw of the vehicle satisfying the correction trigger condition, a correction prompt is displayed on the cruise interface, and the recommended location corresponding to the map marker is maintained. The correction prompt is used to guide the vehicle to travel along the route with the recommended location as the destination.
6. The method according to claim 1, further comprising: In response to the vehicle arriving at the recommended location and the waiting time without receiving an order reaching a first time threshold, a waiting prompt is displayed on the cruise interface. The waiting prompt indicates to continue waiting for an order at the recommended location, and the waiting prompt is determined based on at least one of the following: queuing information, supply and demand information, or estimated waiting time corresponding to the vehicle's current location.
7. The method according to claim 6, further comprising: In response to the waiting time reaching a second duration threshold, updated location information and setting suggestions are presented in the cruise interface. The second duration threshold is greater than the first duration threshold, and the updated location information and setting suggestions are determined based on a re-performed vehicle dispatch recommendation.
8. The method according to claim 1, further comprising: In response to the recommended location remaining valid and the recommendation criteria used to determine the recommended location remaining unchanged, the location information is maintained in the cruise interface, and repeated recommendation broadcasts for the recommended location are suppressed.
9. The method according to claim 1, further comprising: Receive order preference instructions; The order acceptance preference indication is sent to the server so that the server can determine the recommended location and the recommended status based on the matching result of the order acceptance preference indication and the experience record.
10. A method for vehicle dispatch recommendation, executed by a server, comprising: Acquire experience records and vehicle status information. The experience records are obtained based on historical order data statistics and indicate applicable conditions as well as the order location and order setting status corresponding to the applicable conditions. Candidate locations are determined based at least on the matching results between the state information and the applicable conditions, and the candidate locations satisfy the execution constraints for the vehicle. Based on the matching results and the net value of the candidate locations, a recommended location and a recommended status for the order-taking settings are determined. The recommended location is selected from the candidate locations, and the recommended status is determined based on the recommended location and the status of the order-taking settings indicated by the experience record. The net value is determined based on the expected revenue and displacement costs determined for the vehicle and the corresponding candidate location. as well as Provide location information and setting suggestions to the terminal device, wherein the location information indicates the recommended location and the setting suggestions indicate the recommended status.
11. The method of claim 10, wherein obtaining the experience record comprises: For different orders in the historical order data, the status interval set for the order is matched based on the corresponding order acceptance time. The start time of the status interval is no later than the order acceptance time, and the end time of the status interval is later than the order acceptance time. By associating the order acceptance time, starting position, order tag, and the status of the order acceptance settings indicated by the matching status interval for each order, an order acceptance record is obtained. The order receipt records are aggregated to obtain statistical results of the order receipt locations and the status distribution of the order receipt settings. as well as The order-receiving location, the statistical results, and the applicable conditions are associated and stored to obtain the experience record.
12. The method according to claim 11, wherein the applicable conditions include spatiotemporal conditions, the status information includes the current location of the vehicle and the current time period, and the aggregation processing of the order receipt records includes: The order records are segmented based on date type, time period, service evaluation parameters, and order acceptance method category, and the distribution of order acceptance location and order acceptance setting status is statistically analyzed for the obtained segments. A composite key is constructed based on the city identifier, the date type, the time period, the service evaluation parameters, and the order acceptance method category, and the experience record corresponding to the segment is associated with the composite key for storage. And when making vehicle dispatch recommendations, a search key is constructed based on the city identifier corresponding to the current location indicated by the status information, as well as the current date type, time period, service evaluation parameters, and order acceptance method category, and the experience record matching the search key is obtained.
13. The method according to claim 11, wherein the historical order data is obtained in the following manner: Based on city and service evaluation parameters, drivers were divided into sample groups; For the drivers in the sample group, the unit online income, number of valid orders, and completeness of the on / off status records are determined respectively. The historical order-taking data corresponding to drivers whose unit online income reaches the income percentile threshold, whose number of valid orders reaches the quantity threshold, and whose completeness reaches the completeness threshold are selected as the historical order-taking data. The selected historical order data is assigned an order-taking method category according to a preset rule. The preset rule indicates the correspondence between the order-taking behavior statistics and the order-taking method category. The order-taking behavior statistics include the number of orders per unit time or the order distance distribution.
14. The method according to claim 11, wherein the aggregation processing of the order receipt records includes: The latitude and longitude of the starting point are quantized with a predetermined precision to obtain grid coordinates; Construct source location identifiers based on the grid coordinates and city identifiers; as well as Based on the source location identifier, the order receipt records are spatially aggregated to determine the order receipt location.
15. The method of claim 14, wherein spatial aggregation of the order receipt records based on the source location identifier comprises: Based on the source location identifier, the source location and the number of orders corresponding to the source location are determined, and the source location whose number of orders exceeds the seed quantity threshold is determined as the core location; According to the order quantity in descending order, the core locations that have not yet been assigned are used as aggregation centers. Other core locations that are not yet assigned and whose distance from the aggregation center is no greater than a first distance threshold are assigned to the aggregation center to form the order receiving locations. The aggregation center and the core locations assigned to it are marked as assigned. During the selection of the aggregation center, the coverage ratio is determined based on the cumulative number of orders corresponding to the core locations used to form the order receiving locations and the total number of valid orders in the corresponding cities. When the coverage ratio reaches the coverage threshold, the selection of new aggregation centers is stopped. as well as For the source location that has not yet been assigned, determine the order-taking location that is closest to the source location. If the distance is not greater than a second distance threshold, assign the source location to the nearest order-taking location. If the distance is greater than the second distance threshold, exclude the source location from the order-taking locations used to form hotspot experience. The second distance threshold is greater than the first distance threshold.
16. The method of claim 10, wherein determining the candidate location comprises: Obtain locations to be filtered, including the order-accepting locations selected based on the matching results and locations determined based on regional supply and demand data; The locations to be screened from different sources are converted into candidate information in a unified format, the candidate information including location identifier, name, coordinates and source information; Based on the candidate information, the integrity of the name and coordinates is verified, and the feasibility of the locations to be screened is verified according to road accessibility, traffic restrictions, displacement distance restrictions and data validity period. as well as The locations to be screened that pass the integrity check and the executability check are determined as candidate locations, and the candidate locations constitute a candidate whitelist.
17. The method of claim 10, wherein determining the displacement cost comprises: Obtain the path distance and estimated travel time between the candidate location and the vehicle's current location; Based on the path distance and the vehicle's energy cost and depreciation cost parameters, determine the energy cost and depreciation cost; The time cost is determined based on the estimated travel time and the historical revenue per unit of time associated with the vehicle. as well as The displacement cost is determined based on the energy cost, the depreciation cost, and the time cost.
18. The method of claim 16, wherein determining the recommended location comprises: Generate structured input, which includes the candidate whitelist, the net value, the experience matching score, the fusion rule, and the output constraint. The experience matching score represents the matching between the experience data of the candidate location and the state information. The fusion rule indicates the scoring items and the weights of the scoring items. The output constraint limits the selection of one candidate location from the candidate whitelist as the recommended location. The structured input is processed using a language model to obtain a recommendation result, which is determined based on the fusion rules and the output constraints. as well as Based on the recommendation results, the recommended locations and recommendations are determined.
19. The method according to claim 18, wherein the fusion rule is used to determine candidate evaluation results based on the scoring items and the weights corresponding to the scoring items, wherein the scoring items include a value score determined based on the net value and the experience matching score; The method further includes: In response to missing data in the scoring items, a corresponding missing flag is set, the score corresponding to the missing data is set to zero, and the weight corresponding to the scoring item remains unchanged. The missing data includes the absence of matching experience records for the candidate location.
20. The method of claim 18, further comprising: Perform a deterministic validation on the recommendation results, the deterministic validation including: The fusion evaluation results corresponding to the recommendation results are reviewed based on the fusion rules, and the candidate locations indicated by the recommendation results are verified according to the review results to see if they meet the selection conditions specified by the fusion rules; and the candidate locations are verified to see if they belong to the candidate whitelist. Verify whether the recommended description, the location information, and the recommended result correspond to the same name and coordinates; and Verify whether the setting state pointed to by the adjustment operation corresponding to the setting suggestion is consistent with the recommended state determined according to the preset rules, and whether the adjustment operation meets the operation permissions. In response to the successful deterministic verification, the location information and setting suggestions corresponding to the recommendation result are provided to the terminal device; and In response to the failure of the deterministic verification, the navigation operation associated with the recommendation result and the adjustment operation of the order acceptance settings are blocked, and a status description generated based on the status information is provided to the terminal device.
21. The method of claim 10, wherein determining the recommended state comprises: Obtain the order tag distribution corresponding to the recommended location and the status distribution in the experience records associated with the spatiotemporal conditions corresponding to the recommended location, wherein the status distribution indicates the distribution of the order acceptance settings status; as well as The recommended status is determined based on the order tag distribution, the status distribution, the current status of the order acceptance settings, and the operation permissions.
22. The method of claim 21, further comprising: Receive a confirmation instruction from the terminal device regarding the setting suggestion; Based on the confirmation instruction, the conditional action is stored in the session state. The conditional action is associated with the recommended location and the recommended state corresponding to the setting suggestion. The arrival of the vehicle at the recommended location is the trigger condition, and the execution content is to adjust the order acceptance setting to the recommended state. In response to receiving an arrival event indicating that the vehicle has arrived at the recommended location, the conditional action corresponding to the recommended location is retrieved from the session state and executed. as well as In response to receiving an execution receipt, a successful adjustment message is provided to the terminal device, wherein the execution receipt indicates that the order acceptance settings have been adjusted to the recommended state.
23. The method of claim 10, further comprising: Obtain event information indicating the vehicle's cruise event, and update the session state based on the event information. The session state at least indicates the recommended location, navigation status, waiting status, and the status of the order acceptance settings. In response to the vehicle arriving at the recommended location, the arrival time is recorded, and the waiting time for orders that have not been received is determined based on the arrival time; In response to the waiting time reaching a first time threshold, the terminal device is provided with a waiting description determined based on the supply and demand information of the area where the vehicle is located; as well as In response to the waiting time reaching a second time threshold, or the net value of the recommended location decreasing and meeting preset re-recommendation conditions, the candidate location is reacquired and the recommended location and recommendation status are determined, wherein the second time threshold is greater than the first time threshold.
24. The method of claim 20, further comprising: Receive feedback results corresponding to the recommendation results, wherein the feedback results include at least one of the following: execution receipt of setting adjustment, arrival result, order acceptance result, and order completion result; The session status is updated based on the feedback result, and at least one of the following is determined based on the feedback result and the corresponding event time, location record or order data: waiting time, empty driving distance or order revenue. as well as The determined waiting time, empty driving distance, or order revenue is recorded as an effect sample. The effect sample is associated with the experience record used to determine the recommended location and is used to update the effect statistics of the experience record.
25. The method of claim 18, further comprising: The source information is associated with historical statistical evidence and real-time status evidence, respectively, and the historical statistical evidence and the real-time status evidence are used as different fields of the structured input; as well as The recommendation description is generated based on the historical statistical evidence and the real-time status evidence, and the recommendation description distinguishes the sources of the historical statistical evidence and the real-time status evidence.
26. An interactive device for vehicle dispatch recommendation, installed on a terminal device, comprising: The first presentation module is configured to present location information and setting suggestions in the cruise interface. The location information indicates recommended locations, and the setting suggestions indicate the recommended status of the order acceptance settings. The recommended status is associated with the recommended locations. The receiving module is configured to receive a confirmation indication for the setting suggestion; The triggering module is configured to, based on the confirmation indication, trigger the adjustment of the order acceptance settings to the recommended state in response to the vehicle arriving at the recommended location; as well as The second presentation module is configured to display a successful adjustment message in the navigation interface in response to receiving an execution receipt, wherein the execution receipt indicates that the order acceptance settings have been adjusted to the recommended state.
27. A vehicle dispatch recommendation device, installed on a server, comprising: The acquisition module is configured to acquire experience records and vehicle status information. The experience records are obtained based on historical order data statistics and indicate applicable conditions and the order location and order setting status corresponding to the applicable conditions. The candidate determination module is configured to determine candidate locations based at least on the matching result of the state information and the applicable conditions, wherein the candidate locations satisfy the execution constraints for the vehicle; The recommendation determination module is configured to determine a recommended location and a recommendation status of the order acceptance setting based on the matching results and the net value of the candidate locations. The recommended location is selected from the candidate locations. The recommendation status is determined based on the recommended location and the status of the order acceptance setting indicated by the experience record. The net value is determined based on the expected revenue and displacement cost determined for the vehicle and the corresponding candidate location. as well as A module is configured to provide location information and setting suggestions to a terminal device, wherein the location information indicates the recommended location and the setting suggestions indicate the recommended status.
28. An electronic device, comprising: At least one processor; as well as At least one memory coupled to the at least one processor and storing instructions for execution by the at least one processor, the instructions causing the electronic device to perform the method according to any one of claims 1 to 25 when executed by the at least one processor.
29. A computer-readable storage medium having a computer program stored thereon, the computer program being executable by a processor to implement the method according to any one of claims 1 to 25.
30. A computer program product tangibly stored in a computer storage medium and comprising computer-executable instructions that, when executed by a processor, cause the processor to perform the method according to any one of claims 1 to 25.
