A vehicle rental operation management method, device and medium of an internet platform

Abstract

The application discloses an Internet platform vehicle leasing operation management method, device and medium, and relates to the technical field of computers. The method comprises the following steps: obtaining the target and constraint condition of a multi-target optimization algorithm; based on operation data, vehicle state data and market environment data, a multi-target optimization algorithm is used to configure a target optimization operation management process, which comprises sequentially generating a candidate vehicle scheduling scheme, a candidate dynamic pricing strategy and a candidate user service strategy, and establishing a two-way feedback mechanism among the three to iteratively optimize; and performing vehicle leasing operation management according to the target optimization operation management process. Through the three-level progressive optimization architecture and the two-way feedback mechanism, the application realizes the collaborative optimization of vehicle scheduling, dynamic pricing and user service, solves the technical problems of single-point optimization, static decision and credit disconnection in the prior art, and improves the operation efficiency, income level and user satisfaction of the vehicle leasing business.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method, equipment and medium for vehicle rental operation management on an internet platform. Background Technology

[0002] With the rapid development of the sharing economy, the car rental industry has undergone a transformation from traditional offline rental to internet-based and intelligent models. Currently, there are various internet-based car rental platforms on the market, with main technical solutions including C2C sharing models, B2C direct operation models, and time-sharing rental models.

[0003] In modern vehicle rental operations, the configuration of operational management processes directly impacts the efficiency, profitability, and user experience of the rental business. Since vehicle rental operations involve multiple interconnected aspects such as vehicle dispatching, pricing strategies, and customer service, how to plan a scientific and reasonable operational management process has become a key focus for the vehicle rental industry.

[0004] However, existing vehicle rental operation management technologies have the following technical shortcomings: Single-point optimization, lack of collaboration: Existing technologies mostly focus on optimizing a single aspect of vehicle scheduling or dynamic pricing. Some solutions only focus on optimizing vehicle scheduling, achieving coordinated vehicle scheduling through historical data analysis, but fail to consider the linkage between pricing strategies and user services; other solutions combine dynamic pricing and scheduling, but the two are unidirectionally dependent, lacking a two-way feedback mechanism. Independent decision-making at each stage leads to information silos, making it impossible to achieve global optimization.

[0005] Static decision-making makes it difficult to adapt to change: Existing operational management processes are mostly statically configured, unable to be dynamically adjusted according to real-time vehicle status and changes in the market environment. When unexpected situations such as vehicle malfunctions or fluctuations in demand occur, it is difficult to respond quickly.

[0006] Credit assessment is disconnected from operations: Existing user credit assessment technology is mainly used for pre-rental risk control, and is independent of real-time operational decisions (vehicle dispatch, dynamic pricing), failing to achieve dynamic adjustment of differentiated service strategies based on credit rating.

[0007] Lack of closed-loop optimization mechanism: Existing technologies lack a closed-loop feedback mechanism from service execution to strategy optimization, making it impossible to continuously improve operation and management strategies based on user feedback.

[0008] Therefore, there is an urgent need for a vehicle rental operation management method that can achieve multi-stage collaborative optimization, dynamically adapt to environmental changes, and deeply integrate credit assessment with operations. Summary of the Invention

[0009] In view of the aforementioned existing problems, the present invention is proposed.

[0010] Therefore, the present invention provides a vehicle rental operation management method, equipment and medium for an Internet platform to solve one or more of the above-mentioned technical problems.

[0011] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, embodiments of this application provide a vehicle rental operation management method for an internet platform, applied to the process of operating and managing a vehicle rental business on an internet platform, wherein the vehicle rental business involves multiple rental vehicles and multiple users; the method includes: When configuring the vehicle rental operation management process, obtain the objectives and constraints set for the multi-objective optimization algorithm. The objectives include improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction. The constraints include vehicle resource constraints, user demand constraints, operational safety requirements, service quality requirements, and revenue target requirements. Based on the operational data of the current operating scenario, the current vehicle status data of the rental vehicles, and the current market environment data, a target-optimized operation management process is configured using a multi-objective optimization algorithm. The target-optimized operation management process includes: a vehicle intelligent dispatch strategy generated for currently undispatched vehicles, a dynamic pricing strategy generated for the vehicle intelligent dispatch strategy, and a user service strategy generated for the dynamic pricing strategy. Based on the objectives, optimize the operation and management process to control the vehicle rental operation and management of the internet platform.

[0012] In one possible implementation, the method further includes: Detect whether at least one of multiple triggering events has occurred; the multiple triggering events include: receiving an operation management configuration instruction, changes in the vehicle status of the leased vehicle meeting the set vehicle status change threshold, and changes in the market environment exceeding the set market environment change threshold; If this occurs, it is determined that the vehicle rental operation management process needs to be configured.

[0013] In one possible implementation, based on operational data from the current operational scenario, current vehicle status data of the leased vehicles, and current market environment data, a multi-objective optimization algorithm is used to configure a target-optimized operational management process, including: For vehicles that are not currently dispatched, based on operational data, current vehicle status data, current market environment data, and constraints, a spatiotemporal graph neural network prediction model is used to predict the vehicle demand in each region for future periods, generate at least one candidate vehicle dispatching scheme that meets the set vehicle utilization rate requirements, and determine the dispatching scheme score corresponding to each candidate vehicle dispatching scheme. For each candidate vehicle scheduling scheme, based on a multi-agent game model, the revenue level and market competition status under different pricing strategies are simulated to generate at least one candidate pricing strategy that meets the set revenue target requirements, and the pricing strategy score corresponding to each candidate pricing strategy is determined. For each candidate pricing strategy, based on the user's dynamic credit profile, generate at least one candidate user service strategy that meets the set user satisfaction requirements, and determine the service strategy score corresponding to each candidate user service strategy. Generate at least one candidate operation management process, each candidate operation management process including a specific candidate vehicle dispatching scheme, a specific candidate pricing strategy corresponding to the specific candidate vehicle dispatching scheme, and a specific candidate user service strategy corresponding to the specific candidate pricing strategy. Based on the scheduling scheme score corresponding to a specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to a specific candidate pricing strategy, and the service strategy score corresponding to a specific candidate user service strategy, the process score corresponding to each candidate operation management process is determined, and based on the process score, the target preferred operation management process is selected from at least one candidate operation management process.

[0014] In one possible implementation, determining the scheduling scheme score for each candidate vehicle scheduling scheme includes: For each candidate vehicle dispatching scheme, historical order data is analyzed to comprehensively evaluate the average daily number of vehicle orders, vehicle empty running rate, and vehicle idle time to determine the vehicle utilization score. For each candidate vehicle dispatching scheme, the dispatching cost score is determined by calculating the dispatching distance, dispatching energy consumption, and dispatching manpower cost; Obtain the pre-set vehicle utilization rate score weights and scheduling cost score weights; The scheduling plan score is calculated by weighting the vehicle utilization rate score, the scheduling cost score, the weight of the vehicle utilization rate score, and the weight of the scheduling cost score.

[0015] In one possible implementation, determining the pricing strategy score for each candidate pricing strategy includes: For each candidate pricing strategy, by analyzing historical pricing data, simulating the execution of dynamic pricing strategies, predicting expected returns, and determining return scores; For each candidate pricing strategy, by analyzing market competition data, simulating the execution of dynamic pricing strategies, predicting changes in market share and user churn rate, and determining the market competitiveness score; Obtain the pre-set profit rating weights and market competitiveness rating weights; The pricing strategy score is calculated by weighting the revenue score, market competitiveness score, revenue score weight, and market competitiveness score weight.

[0016] In one possible implementation, determining the service policy score corresponding to each candidate user service policy includes: For each candidate user service strategy, based on the user's dynamic credit profile, the response time, service accuracy, and complaint handling efficiency are evaluated to determine the service quality score; For each candidate user service strategy, based on the user's dynamic credit profile, predict the user repurchase rate, user satisfaction, and user recommendation intention, and determine the user retention score. In one possible implementation, pre-set service quality score weights and user retention score weights are obtained; The service strategy score is calculated by weighting the service quality score, user retention score, service quality score weight, and user retention score weight.

[0017] In one possible implementation, the process score for each candidate operation management process is determined based on the scheduling scheme score corresponding to a specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to a specific candidate pricing strategy, and the service strategy score corresponding to a specific candidate user service strategy. Obtain the pre-set scoring weights for scheduling schemes, pricing strategies, and service strategies; The process score is calculated based on the scheduling scheme score corresponding to a specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to a specific candidate pricing strategy, the service strategy score corresponding to a specific candidate user service strategy, the scheduling scheme score weight, the pricing strategy score weight, and the service strategy score weight.

[0018] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor implements the method provided in any embodiment of this application when executing the computer program.

[0019] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method provided in any embodiment of this application.

[0020] The beneficial effects of this invention are as follows: The technical solution of this application can flexibly configure a target-optimized operation management process based on operational data, vehicle status data, and market environment data of the current operating scenario to guide the internet platform in executing vehicle rental operation management. In this case, when vehicle status or market environment changes, the technical solution of this application can dynamically adjust and configure a new target-optimized operation management process in real time based on current data, thereby enabling the target-optimized operation management process to adapt to changes in the environment or status. At this time, guiding the internet platform to execute operation management according to this target-optimized operation management process ensures the smooth operation and security of the vehicle rental business.

[0021] Furthermore, since the technical solution of this application employs a multi-objective optimization algorithm to configure the target-optimized operation management process, the set optimization objectives include improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction. Relevant constraints cover vehicle resource constraints, user demand constraints, operational safety requirements, service quality requirements, and revenue target requirements. Therefore, each configured target-optimized operation management process achieves the current optimal solution for improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction. Under these conditions, the entire vehicle rental operation can be guaranteed to have high operational efficiency and significantly reduce operating costs while ensuring operational safety and user satisfaction, thereby improving the overall efficiency of the vehicle rental business.

[0022] More importantly, the technical solution of this application breaks down information silos in existing technologies by establishing a two-way feedback mechanism among candidate vehicle scheduling schemes, candidate dynamic pricing strategies, and candidate user service strategies, achieving coordinated optimization of the three aspects of scheduling, pricing, and service. By feeding user feedback data back to the demand forecasting model to correct parameters, and feeding user satisfaction data back to the pricing strategy model to correct parameters, a closed loop is formed from strategy formulation to execution feedback and then to strategy optimization, enabling continuous iterative optimization of the operation and management process. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This application illustrates a vehicle rental operation management method for an internet platform provided in an embodiment of the present application; Figure 2 This application shows a structural block diagram of a vehicle rental operation management system for an internet platform according to an embodiment of the present application; Figure 3 A schematic diagram of a three-level progressive optimization architecture according to an embodiment of this application is shown; Figure 4 A schematic diagram of a bidirectional feedback mechanism according to an embodiment of this application is shown; Figure 5 A block diagram of an electronic device according to an embodiment of this application is shown. Detailed Implementation

[0025] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0026] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0027] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0028] Figure 1 The flowchart illustrates a vehicle rental operation management method 100 for an internet platform provided in this embodiment of the application. This vehicle rental operation management method is applied to the process of operating and managing vehicle rental business on an internet platform, and the method may include the following steps: Step S101: When configuring the vehicle rental operation management process, obtain the objectives and constraints set for the multi-objective optimization algorithm. The objectives include improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction. The constraints include vehicle resource constraints, user demand constraints, operational safety requirements, service quality requirements, and revenue target requirements.

[0029] This step specifically involves: collecting user behavior data, credit data, and location data through user terminals; collecting vehicle status data, location data, and driving behavior data through in-vehicle intelligent terminals; and collecting traffic data, weather data, and POI data through third-party interfaces. The collected data is then cleaned, standardized, and spatiotemporally correlated and fused to construct a unified spatiotemporal data view. Based on this view, a spatiotemporal graph neural network prediction model is used to predict future vehicle demand in various regions. Combined with current vehicle status data, a multi-objective optimization algorithm is used to generate candidate vehicle scheduling schemes. Based on these schemes, a multi-agent game model is used to simulate different pricing strategies, generating candidate dynamic pricing strategies. Based on these pricing strategies and user credit profiles, candidate user service strategies are generated. A two-way feedback mechanism is established between the three strategies, and the optimal operational management process is determined through iterative optimization.

[0030] Step S102: Based on the operational data of the current operating scenario, the current vehicle status data of the rental vehicles, and the current market environment data, a target-optimized operation management process is configured using a multi-objective optimization algorithm. The target-optimized operation management process includes the vehicle intelligent dispatch strategy corresponding to the rental vehicles, the dynamic pricing strategy configured for the vehicle intelligent dispatch strategy, and the user service strategy configured for the dynamic pricing strategy.

[0031] This step specifically involves: Utilizing operational data from the current operational scenario, current vehicle status data of the leased vehicles, and current market environment data, a multi-objective optimization algorithm is used to configure a target-optimized operational management process. This process employs a three-tiered progressive optimization architecture: The first tier uses a spatiotemporal graph neural network prediction model to predict future vehicle demand in various regions, generating candidate vehicle scheduling schemes that meet set vehicle utilization requirements, and determining the scheduling scheme score based on a weighted average of vehicle utilization and scheduling cost scores. The second tier uses a multi-agent game model to simulate revenue levels and market competition under different pricing strategies, generating candidate dynamic pricing strategies that meet set revenue targets, and determining the pricing strategy score based on a weighted average of revenue and market competitiveness scores. The third tier uses dynamic user credit profiles to generate candidate user service strategies that meet set user satisfaction requirements, and determines the service strategy score based on a weighted average of service quality and user retention scores. A two-way feedback mechanism is established between the three strategies, feeding user feedback data on pricing strategies back to the demand prediction model's correction parameters, and feeding user service strategy satisfaction data back to the pricing strategy model's correction parameters, thus determining the target-optimized operational management process through iterative optimization.

[0032] Step S103: Based on the target optimization operation management process, control the Internet platform to execute vehicle rental operation management.

[0033] The vehicle rental operation management method provided in this application embodiment can flexibly configure a target-optimized operation management process based on the current operation scenario's operational data, current vehicle status data, and current market environment data to guide the internet platform in executing vehicle rental operation management. In this case, when vehicle status or the market environment changes, the vehicle rental operation management method provided in this application embodiment can, in real time, based on current data ("current data" includes: historical orders, user behavior, vehicle usage, financial and market competition data, etc.; vehicle status data such as vehicle location, battery level, and health; market environment data such as traffic, weather, holidays, and competitor dynamics), re-collect current data, dynamically adjust and configure a new target-optimized operation management process using a multi-objective optimization algorithm: re-predict future vehicle demand in each region and generate candidate vehicle dispatching schemes; simulate pricing strategies based on the new dispatching schemes and generate candidate dynamic pricing strategies; generate candidate user service strategies based on the new pricing strategies; establish a new two-way feedback mechanism, iteratively optimize and determine the new target-optimized operation management process to guide the internet platform in executing operation management, ensuring that the vehicle rental business adapts to environmental changes and guarantees operational safety.

[0034] Furthermore, since the vehicle rental operation management method provided in this embodiment employs a multi-objective optimization algorithm to configure the target-optimized operation management process, the set optimization objectives include improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction. The relevant constraints cover vehicle resource constraints, user demand constraints, operational safety requirements, service quality requirements, and revenue target requirements. Therefore, each configured target-optimized operation management process achieves the current optimal solution for improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction. Under these circumstances, it can be guaranteed that the entire vehicle rental operation has high operational efficiency and significantly reduces operating costs while ensuring operational safety and user satisfaction, thereby improving the overall efficiency of the vehicle rental business.

[0035] It should be noted that the executing entity of the vehicle rental operation management method provided in this application embodiment can be a hardware device with vehicle rental operation management functions, such as a server or terminal device. The server can be a cloud server or an edge computing server. The terminal device can be an in-vehicle terminal device or a handheld terminal device. In this application embodiment, no specific limitation is made on the executing entity corresponding to the vehicle rental operation management method provided in this application embodiment.

[0036] Vehicle rental services refer to the business of providing travel services to users through internet platforms by renting vehicles. Common rental vehicles include traditional gasoline-powered vehicles, pure electric vehicles, and hybrid vehicles. This application does not specifically limit the number or type of rental vehicles.

[0037] The vehicle rental business in this application involves multiple rental vehicles and multiple users. In one example, a city-wide car-sharing platform operates 5,000 vehicles, covering 500 operational grids in the main urban area, serving 1 million registered users. In another example, an enterprise-level long-term rental platform manages 10,000 vehicles, serving 500 corporate clients.

[0038] Operational safety includes both the safety of the vehicle during operation and the safety of the user during vehicle use.

[0039] When setting objectives, in addition to improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction, objectives can be further set according to business requirements, such as reducing carbon emissions or increasing brand awareness. That is to say, the objectives in this embodiment are not specifically limited. Correspondingly, when setting constraints, in addition to constraints including vehicle resource constraints, user demand constraints, operational safety requirements, service quality requirements, and revenue target requirements, constraints can also be further set according to business requirements, such as policy and regulatory restrictions and social responsibility requirements.

[0040] In this embodiment of the application, the operational data of the current operating scenario includes, but is not limited to: historical order data, user behavior data, vehicle usage data, financial data, market competition data, etc.

[0041] Vehicle status includes, but is not limited to: vehicle location, vehicle battery or fuel level, vehicle health, vehicle cleanliness, and vehicle availability.

[0042] Market environment includes, but is not limited to: traffic conditions, weather conditions, holiday arrangements, information on major events, and competitor activities.

[0043] In this embodiment, multiple trigger events can be pre-set. Then, it is detected whether at least one of the multiple trigger events has occurred, and when at least one of the multiple trigger events is detected, it is determined that the vehicle rental operation management process needs to be configured.

[0044] Three types of trigger events are pre-defined: first, operation management configuration commands, manually triggered by operation managers through the management interface, applicable to initial configuration or emergencies; second, vehicle status change thresholds, including vehicle health score falling below a set threshold, battery or fuel level falling below a safety threshold, and vehicle malfunction alarms, monitored in real time by onboard sensors; and third, market environment change thresholds, including sudden weather changes, severe traffic congestion, and competitors launching large-scale promotional activities, obtained in real time through third-party data interfaces. The system continuously monitors these trigger events. When at least one event occurs, it determines that the vehicle rental operation management process needs to be configured, automatically initiating a multi-objective optimization algorithm to reconfigure the target-optimized operation management process, achieving dynamic adaptation to environmental changes.

[0045] Multiple triggering events include receiving operation management configuration instructions, changes in the rental vehicle's status meeting set vehicle status change thresholds, and changes in the market environment exceeding set market environment change thresholds. Threshold settings are based on: vehicle safety operation standards and technical specifications, where the vehicle health score threshold of 60 points is based on vehicle technical condition level classifications (Excellent ≥80, Good 60-79, Average 40-59), the battery threshold of 20% is based on the electric vehicle industry's low battery warning standard, and the fault threshold is based on OBD diagnostic system standards; weather warning levels, traffic condition indices, and competitor monitoring data, where the weather threshold is based on rainstorm / snowstorm / extreme weather warnings issued by meteorological departments, the traffic threshold is based on the severe congestion standard of a congestion index >8, and the competitor threshold is based on the market competitiveness judgment standard of discounts >30%.

[0046] When initial operation management process configuration is required, operation management configuration instructions can be generated in response to the actions of relevant operation management personnel, thereby triggering the initial operation management process configuration. In addition, in the event of emergencies or at the completion of each key operation cycle, operation management configuration instructions can also be generated in response to the actions of relevant operation management personnel, thereby triggering the reconfiguration of the operation management process.

[0047] In this embodiment of the application, configuring the operation management process includes configuring the operation management process for the first time and reconfiguring the operation management process.

[0048] In this embodiment, a change in the vehicle's status that meets a set threshold can refer to a vehicle health score falling below a set threshold, a battery or fuel level falling below a safety threshold, or a vehicle malfunction alarm. These changes indicate a significant change in the vehicle's status that may affect its safety and availability.

[0049] The aforementioned thresholds are preset and stored in the system database through the following methods: first, based on national and industry technical standards (such as GB / T 32960 "Technical Specifications for Remote Service and Management System of Electric Vehicles" and JT / T 617 "Technical Grade Classification and Evaluation Requirements for Road Transport Vehicles"); second, based on the technical parameters and safety operation manuals provided by the vehicle manufacturers; and third, dynamically adjusted and set by operation and management personnel through the configuration interface of the management control module based on actual operation experience and historical fault data analysis.

[0050] The data acquisition module collects vehicle sensor data in real time through the onboard intelligent terminal, including SOC (State of Charge) reported by the battery management system, fault codes (PIDs) reported by the engine control unit, and a comprehensive score calculated by the vehicle health assessment model. When any indicator reaches the above-mentioned set threshold, the system automatically triggers a reconfiguration of the vehicle rental operation management process to ensure operational safety.

[0051] In this embodiment of the application, a market environment change exceeding a set market environment change threshold may refer to a change in weather from sunny to rainy, a change in traffic conditions from smooth to severe congestion, or a competitor's platform launching a large-scale promotional activity.

[0052] In one possible implementation, when configuring a target-optimized operation management process using a multi-objective optimization algorithm based on operational data of the current operational scenario, current vehicle status data of rental vehicles, and current market environment data, the following approach can be taken: First, for currently unscheduled vehicles, a spatiotemporal graph neural network prediction model can be used to predict the future vehicle demand in each region. This generates at least one candidate vehicle scheduling scheme that meets the set vehicle utilization rate requirements, and determines the scheduling scheme score corresponding to each candidate vehicle scheduling scheme. The spatiotemporal graph neural network prediction model takes the spatiotemporal distribution of historical order data as input, combines the current vehicle location, regional POI characteristics, and weather factors, and outputs the predicted vehicle demand value for each operational grid in the next 1-4 hours. The vehicle utilization rate requirements include an average daily number of orders per vehicle ≥ 3, a vehicle empty-running rate ≤ 15%, and a vehicle idle time ≤ 4 hours. The scheduling scheme score is obtained by weighted calculation of the vehicle utilization rate score and the scheduling cost score. The vehicle utilization rate score is based on a comprehensive evaluation of the average daily number of orders per vehicle, empty-running rate, and idle time, while the scheduling cost score is calculated based on scheduling distance, scheduling energy consumption, and scheduling manpower costs.

[0053] For each candidate vehicle dispatching scheme, at least one candidate dynamic pricing strategy that meets the set revenue target is generated based on a multi-agent game model, and the pricing strategy score corresponding to each candidate pricing strategy is determined. The multi-agent game model simulates the price game process between this platform and competing platforms. Taking the regional supply-demand ratio under each candidate vehicle dispatching scheme as input, the output is a dynamic pricing strategy that maximizes the platform's revenue and stabilizes its market share. The revenue target requirements include an expected rate of return ≥ 15% and a price competitiveness index ≥ 0.8. The pricing strategy score is obtained by weighted calculation of the revenue score and the market competitiveness score, where the revenue score is determined based on the expected revenue level, and the market competitiveness score is determined based on the predicted changes in market share and the predicted user churn rate.

[0054] Next, for each candidate dynamic pricing strategy, at least one candidate user service strategy that meets the set user satisfaction requirements is generated based on the user's dynamic credit profile, and the service strategy score corresponding to each candidate user service strategy is determined. The user dynamic credit profile calculates a credit score by comprehensively considering the user's historical performance records, real-time behavioral data, and third-party credit data. Users with a credit score ≥80 are considered high-credit users, 60-79 are medium-credit users, and <60 are low-credit users. Differentiated service strategies are generated for users with different credit levels, including response time (≤5 minutes for high-credit users, ≤10 minutes for medium-credit users, and ≤15 minutes for low-credit users), deposit reduction ratio, and priority car use rights. User satisfaction requirements include a user satisfaction score ≥4.5, a user complaint rate ≤2%, and a user repurchase rate ≥60%. The service strategy score is obtained by weighted calculation of the service quality score and the user retention score, where the service quality score is based on the evaluation of response time, service accuracy, and complaint handling efficiency, and the user retention score is determined based on the user repurchase rate, satisfaction, and recommendation intention prediction.

[0055] Then, a two-way feedback mechanism is established between candidate vehicle scheduling schemes, candidate dynamic pricing strategies, and candidate user service strategies. Based on the iterative optimization of the two-way feedback mechanism, the target optimal operation management process is determined. The two-way feedback mechanism includes: forward transmission, which is the hierarchical progression of candidate vehicle scheduling schemes → candidate dynamic pricing strategies → candidate user service strategies; and reverse feedback, which feeds back user feedback data (including actual order volume, price sensitivity, and reasons for order cancellation) after the implementation of candidate dynamic pricing strategies to the spatiotemporal graph neural network prediction model to correct demand prediction parameters; and feeds back user satisfaction data (including service evaluation, complaint content, and repurchase behavior) after the implementation of candidate user service strategies to the multi-agent game model to correct pricing strategy parameters. The iterative optimization process involves 3-5 rounds of two-way feedback iterations, ensuring that the scores of the three strategies all reach preset thresholds or converge, and finally selecting the candidate operation management process with the highest process score as the target optimal operation management process.

[0056] In one example, a city is divided into 100 operational grids. One candidate vehicle dispatching scheme is to dispatch 10 vehicles to grid A and 5 vehicles to grid B. Another candidate vehicle dispatching scheme is to dispatch 8 vehicles to grid A and 7 vehicles to grid C.

[0057] In one possible implementation, the scheduling scheme score corresponding to each candidate vehicle scheduling scheme is determined as follows: First, determine the vehicle utilization rate score and scheduling cost score corresponding to each candidate vehicle scheduling scheme. Then, obtain the pre-set weights for the vehicle utilization rate score and scheduling cost score. Finally, calculate the scheduling scheme score based on the vehicle utilization rate score, scheduling cost score, vehicle utilization rate score weight, and scheduling cost score weight.

[0058] In this embodiment, the dynamic pricing strategy refers to differentiated pricing schemes for different regions, time periods, and vehicle models. Specifically, it may include: a base price setting, time-based premiums or discount coefficients, regional adjustment coefficients, and vehicle model price differences.

[0059] In one example, there are three candidate vehicle scheduling schemes: candidate vehicle scheduling scheme 1, candidate vehicle scheduling scheme 2, and candidate vehicle scheduling scheme 3. For candidate vehicle scheduling scheme 1, two candidate dynamic pricing strategies are generated: candidate dynamic pricing strategy 1 and candidate dynamic pricing strategy 2. For candidate vehicle scheduling scheme 2, two candidate dynamic pricing strategies are generated: candidate dynamic pricing strategy 3 and candidate dynamic pricing strategy 4. For candidate vehicle scheduling scheme 3, two candidate dynamic pricing strategies are generated: candidate dynamic pricing strategy 5 and candidate dynamic pricing strategy 6.

[0060] In one possible implementation, the pricing strategy score for each candidate dynamic pricing strategy is determined as follows: First, the revenue score and market competitiveness score for each candidate dynamic pricing strategy are determined. Then, pre-set weights for the revenue score and market competitiveness score are obtained. Finally, the pricing strategy score is calculated based on the revenue score, market competitiveness score, revenue score weights, and market competitiveness score weights.

[0061] The revenue score and market competitiveness score are obtained by analyzing historical pricing data and market competition data, respectively, and simulating the execution of dynamic pricing strategies.

[0062] In one example, there are six candidate dynamic pricing strategies: candidate dynamic pricing strategy 1, candidate dynamic pricing strategy 2, candidate dynamic pricing strategy 3, candidate dynamic pricing strategy 4, candidate dynamic pricing strategy 5, and candidate dynamic pricing strategy 6. Among them, candidate dynamic pricing strategy 1 generates two candidate user service strategies: candidate user service strategy 1 and candidate user service strategy 2. Candidate dynamic pricing strategy 2 generates two candidate user service strategies: candidate user service strategy 3 and candidate user service strategy 4. Candidate dynamic pricing strategy 3 generates two candidate user service strategies: candidate user service strategy 5 and candidate user service strategy 6. Candidate dynamic pricing strategy 4 generates two candidate user service strategies: candidate user service strategy 7 and candidate user service strategy 8. Candidate dynamic pricing strategy 5 generates two candidate user service strategies: candidate user service strategy 9 and candidate user service strategy 10. Candidate dynamic pricing strategy 6 generates two candidate user service strategies: candidate user service strategy 11 and candidate user service strategy 12.

[0063] In this scenario, 12 candidate operation management processes are generated. Candidate operation management process 1 includes: Candidate vehicle dispatching scheme 1, Candidate dynamic pricing strategy 1, and Candidate user service strategy 1. Candidate operation management process 2 includes: Candidate vehicle dispatching scheme 1, Candidate dynamic pricing strategy 1, and Candidate user service strategy 2. Candidate operation management process 3 includes: Candidate vehicle dispatching scheme 1, Candidate dynamic pricing strategy 2, and Candidate user service strategy 3. Candidate operation management process 4 includes: Candidate vehicle dispatching scheme 1, Candidate dynamic pricing strategy 2, and Candidate user service strategy 4. Candidate operation management process 5 includes: Candidate vehicle dispatching scheme 2, Candidate dynamic pricing strategy 3, and Candidate user service strategy 5. Candidate operation management process 6 includes: Candidate vehicle dispatching scheme 2, Candidate dynamic pricing strategy 3, and Candidate user service strategy 6. Candidate operation management process 7 includes: Candidate vehicle dispatching scheme 2, Candidate dynamic pricing strategy 4, and Candidate user service strategy 7. Candidate operation management process 8 includes: Candidate vehicle dispatching scheme 2, Candidate dynamic pricing strategy 4, and Candidate user service strategy 8. Candidate Operations Management Process 9 includes: Candidate Vehicle Dispatch Scheme 3, Candidate Dynamic Pricing Strategy 5, and Candidate User Service Strategy 9. Candidate Operations Management Process 10 includes: Candidate Vehicle Dispatch Scheme 3, Candidate Dynamic Pricing Strategy 5, and Candidate User Service Strategy 10. Candidate Operations Management Process 11 includes: Candidate Vehicle Dispatch Scheme 3, Candidate Dynamic Pricing Strategy 6, and Candidate User Service Strategy 11. Candidate Operations Management Process 12 includes: Candidate Vehicle Dispatch Scheme 3, Candidate Dynamic Pricing Strategy 6, and Candidate User Service Strategy 12.

[0064] In one possible implementation, establishing a two-way feedback mechanism among candidate vehicle scheduling schemes, candidate dynamic pricing strategies, and candidate user service strategies includes: feeding user feedback data of candidate dynamic pricing strategies into a spatiotemporal graph neural network prediction model to correct demand prediction parameters; and feeding user satisfaction data of candidate user service strategies into a multi-agent game model to correct pricing strategy parameters.

[0065] In one possible implementation, when determining the process score for each candidate operation management process based on the scheduling scheme score corresponding to a specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to a specific candidate dynamic pricing strategy, and the service strategy score corresponding to a specific candidate user service strategy, pre-set scheduling scheme score weights, pricing strategy score weights, and service strategy score weights can be obtained first. Then, the process score is calculated based on the scheduling scheme score corresponding to the specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to the specific candidate dynamic pricing strategy, the service strategy score corresponding to the specific candidate user service strategy, the scheduling scheme score weights, the pricing strategy score weights, and the service strategy score weights.

[0066] The weighting method and specific values ​​are as follows: The weights for scheduling scheme scoring, pricing strategy scoring, and service strategy scoring are preset by operations management personnel through the configuration interface of the management control module, or automatically optimized and determined based on historical operational data through the model training module. In standard operational scenarios, the three weights are balanced: scheduling scheme scoring weight is 0.4, pricing strategy scoring weight is 0.35, and service strategy scoring weight is 0.25. This is based on the following: vehicle scheduling is the fundamental guarantee of operations, directly affecting vehicle utilization and operating costs, hence its highest weight; dynamic pricing is the core of revenue, hence its second highest weight; and user service is a long-term competitive advantage, hence its relatively lower weight. In specific operational scenarios, the weights can be dynamically adjusted: for example, during the new market expansion phase, the service strategy scoring weight is increased to 0.4 to prioritize user experience; during periods of revenue pressure, the pricing strategy scoring weight is increased to 0.45 to prioritize revenue targets; and during periods of vehicle shortage, the scheduling scheme scoring weight is increased to 0.5 to prioritize the rational allocation of resources. The weight settings must meet the constraint that the sum of the three is 1.0 to ensure the normalized calculation of the process score.

[0067] The process score calculation formula is as follows: the process score equals the scheduling plan score multiplied by its weight, plus the pricing strategy score multiplied by its weight, plus the service strategy score multiplied by its weight, and finally, the weighted sum of these three scores. This weighted calculation method ensures that the contribution of the three-tiered strategies to the process score matches their respective weight settings, thereby objectively reflecting the overall merits of the candidate operation and management processes.

[0068] In one possible implementation, when selecting the target preferred operation management process from at least one candidate operation management process based on process scoring, the candidate operation management process with the highest process score can be selected as the target preferred operation management process from at least one candidate operation management process.

[0069] In this embodiment, a target optimization operation management process is configured using a multi-objective optimization algorithm based on the operation data of the current operation scenario, the current vehicle status data of the rental vehicles, and the current market environment data. In addition to configuring the vehicle dispatching plan, dynamic pricing strategy, and user service strategy in sequence, it can also configure only the vehicle dispatching plan and dynamic pricing strategy, or only the dynamic pricing strategy and user service strategy.

[0070] Figure 2 The diagram shows a structural block diagram of a vehicle rental operation management system 200 for an internet platform according to an embodiment of this application. The system 200 may include: a data acquisition module 201, a multi-source data fusion engine 202, a demand forecasting engine 203, a scheduling optimization engine 204, a pricing decision engine 205, a credit assessment engine 206, a service strategy engine 207, a feedback control module 208, a business execution module 209, a user interaction module 210, and a management control module 211.

[0071] Specifically, the data acquisition module 201 collects user behavior data, credit data, and location data through user terminals; vehicle status data, location data, and driving behavior data through in-vehicle intelligent terminals; and traffic data, weather data, and POI data through third-party interfaces. The multi-source data fusion engine 202 cleans, standardizes, and spatiotemporally correlates and fuses the collected data to construct a unified spatiotemporal data view. The demand forecasting engine 203, deployed in the data processing module, predicts future vehicle demand in various regions based on the spatiotemporal data view using a spatiotemporal graph neural network prediction model. The scheduling optimization engine 204, also deployed in the data processing module, generates vehicle scheduling plans based on prediction results and current vehicle status using a multi-objective optimization algorithm. The pricing decision engine 205, deployed in the data processing module, generates dynamic pricing strategies based on vehicle scheduling plans and the current market environment using a multi-agent game model. The credit assessment engine 206, deployed in the data processing module, calculates dynamic credit scores based on historical user data and real-time behavior and feeds them back to the scheduling optimization engine and the pricing decision engine. The service strategy engine 207, deployed within the data processing module, generates differentiated user service strategies based on dynamic pricing strategies and user credit scores. The feedback control module 208 establishes a bidirectional data channel between the scheduling optimization engine, pricing decision engine, credit assessment engine, and service strategy engine, enabling real-time linkage and iterative optimization between strategies. The business execution module 209 optimizes operational management processes based on objectives, executing intelligent vehicle scheduling, dynamic pricing, and user services. The user interaction module 210 responds to specific user actions triggered by the user interface, providing vehicle inquiry, reservation, payment, and return services. The management control module 211 generates corresponding specific operational management instructions in response to specific actions triggered by operational management personnel through the designated management interface.

[0072] In one possible implementation, the system 200 may further include a data storage module 212 for storing vehicle sensor data using a time-series database, storing user-vehicle relationship networks using a graph database, and storing business data using a relational database.

[0073] In one possible implementation, the system 200 may further include a model training module 213 for training a spatiotemporal graph neural network prediction model and a multi-agent game model using historical data, optimizing parameters using a stochastic gradient descent optimizer, and using an early stopping mechanism to prevent overfitting.

[0074] The functions of each module in each system of the embodiments of this application can be found in the corresponding description in the above method, and they have corresponding beneficial effects, which will not be repeated here.

[0075] Figure 3A schematic diagram of a three-level progressive optimization architecture 300 according to an embodiment of this application is shown. The architecture 300 includes: First layer: Vehicle scheduling layer 301, which generates candidate vehicle scheduling schemes based on a spatiotemporal graph neural network prediction model; The second layer: pricing strategy layer 302, which generates candidate dynamic pricing strategies based on a multi-agent game model; The third layer: Service strategy layer 303, which generates candidate user service strategies based on dynamic user credit profiles; Two-way feedback 304: Establish a two-way feedback mechanism between the three layers to achieve iterative optimization.

[0076] Figure 4 A schematic diagram of a bidirectional feedback mechanism 400 according to an embodiment of this application is shown. The mechanism 400 includes: Forward transit 401: Candidate vehicle dispatching scheme → Candidate dynamic pricing strategy → Candidate user service strategy; Feedback 402: User feedback data is fed back to the demand forecasting model, and user satisfaction data is fed back to the pricing strategy model.

[0077] Figure 5 This is a block diagram of an electronic device used to implement embodiments of this application. For example... Figure 5 As shown, the electronic device includes a memory 501 and a processor 502. The memory 501 stores a computer program that can run on the processor 502. When the processor 502 executes the computer program, it implements the method described in the above embodiments. The number of memories 501 and processors 502 can be one or more.

[0078] The electronic device also includes a communication interface 503 for communicating with external devices and exchanging and transmitting data.

[0079] If the memory 501, processor 502, and communication interface 503 are implemented independently, they can be interconnected via a bus to communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0080] Optionally, in a specific implementation, if the memory 501, processor 502, and communication interface 503 are integrated on a single chip, then the memory 501, processor 502, and communication interface 503 can communicate with each other through an internal interface.

[0081] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method provided in this application.

[0082] This application also provides a chip, which includes a processor for calling and executing instructions stored in a memory, causing a communication device on which the chip is installed to perform the method provided in this application.

[0083] This application also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, output interface, processor, and memory are connected through an internal connection path. The processor is used to execute code in the memory. When the code is executed, the processor is used to execute the method provided in the application embodiment.

[0084] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0085] Further, optionally, the aforementioned memory may include read-only memory and random access memory, and may also include non-volatile random access memory. The memory may be volatile or non-volatile, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. Many forms of RAM are available by way of example, but not limitation. Examples include Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0086] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0087] In summary, this application addresses the technical problems of existing technologies, such as single-point optimization, static decision-making, credit disconnect, lack of closed loop, single objective, and fixed weights, through a three-tiered progressive optimization architecture, a two-way feedback mechanism, real-time linkage between credit assessment and operational decision-making, dynamic adaptability, scenario-based application of multi-objective optimization algorithms, and dynamic adjustment of scoring weights. This results in improved operational efficiency, increased profitability, enhanced user satisfaction, strengthened risk control, improved adaptability, and more precise strategy matching, significantly improving the overall operational level of the vehicle rental business.

[0088] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for managing vehicle rental operations on an internet platform, characterized in that: The method includes, When configuring the vehicle rental operation management process, obtain the objectives and constraints set for the multi-objective optimization algorithm; the objectives include improving vehicle utilization and reducing operating costs while ensuring operational safety and user satisfaction; the constraints include vehicle resource constraints, user demand constraints, operational safety requirements, service quality requirements, and revenue target requirements; Based on the operational data of the current operating scenario, the current vehicle status data of the rental vehicles, and the current market environment data, the multi-objective optimization algorithm is used to configure a target-optimized operation management process. The target optimization operation management process includes: a vehicle intelligent dispatch strategy generated for currently undispatched vehicles, a dynamic pricing strategy generated for the vehicle intelligent dispatch strategy, and a user service strategy generated for the dynamic pricing strategy. Based on the aforementioned target optimization operation and management process, the internet platform is controlled to perform vehicle rental operation and management.

2. The vehicle rental operation management method according to claim 1, characterized in that: The method further includes: The system detects whether at least one of a variety of triggering events has occurred; the various triggering events include: receiving an operation management configuration instruction, the vehicle status change of the rental vehicle meeting a set vehicle status change threshold, and market environment changes exceeding a set market environment change threshold; If this occurs, it is determined that the vehicle rental operation management process needs to be configured.

3. The vehicle rental operation management method according to claim 1, characterized in that: The operational data for the current operating scenario, the current vehicle status data of the rental vehicles, and the current market environment data, combined with the multi-objective optimization algorithm, are used to configure the target-optimized operation management process, which includes: For vehicles that are not currently dispatched, based on the operational data, the current vehicle status data, the current market environment data, and the constraints, a spatiotemporal graph neural network prediction model is used to predict the vehicle demand in each region for future periods, generate at least one candidate vehicle dispatching scheme that meets the set vehicle utilization rate requirements, and determine the dispatching scheme score corresponding to each candidate vehicle dispatching scheme. For each candidate vehicle scheduling scheme, based on a multi-agent game model, the revenue level and market competition status under different pricing strategies are simulated to generate at least one candidate pricing strategy that meets the set revenue target requirements, and the pricing strategy score corresponding to each candidate pricing strategy is determined. For each candidate pricing strategy, based on the user's dynamic credit profile, generate at least one candidate user service strategy that meets the set user satisfaction requirements, and determine the service strategy score corresponding to each candidate user service strategy. Generate at least one candidate operation management process, each candidate operation management process including a specific candidate vehicle dispatching scheme, a specific candidate pricing strategy corresponding to the specific candidate vehicle dispatching scheme, and a specific candidate user service strategy corresponding to the specific candidate pricing strategy; Based on the scheduling scheme score corresponding to the specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to the specific candidate pricing strategy, and the service strategy score corresponding to the specific candidate user service strategy, the process score corresponding to each candidate operation management process is determined, and based on the process score, the target preferred operation management process is selected from the at least one candidate operation management process.

4. The vehicle rental operation management method according to claim 3, characterized in that: The determination of the scheduling scheme score corresponding to each candidate vehicle scheduling scheme includes: For each candidate vehicle dispatching scheme, historical order data is analyzed to comprehensively evaluate the average daily number of vehicle orders, vehicle empty running rate, and vehicle idle time to determine the vehicle utilization score. For each candidate vehicle dispatching scheme, the dispatching cost score is determined by calculating the dispatching distance, dispatching energy consumption, and dispatching manpower cost; Obtain the pre-set vehicle utilization rate score weights and scheduling cost score weights; The scheduling scheme score is calculated by weighting the vehicle utilization rate score, the scheduling cost score, the vehicle utilization rate score weight, and the scheduling cost score weight.

5. The vehicle rental operation management method according to claim 3, characterized in that: The determination of the pricing strategy score corresponding to each candidate pricing strategy includes: For each candidate pricing strategy, by analyzing historical pricing data, simulating the execution of dynamic pricing strategies, predicting expected returns, and determining return scores; For each candidate pricing strategy, by analyzing market competition data, simulating the execution of dynamic pricing strategies, predicting changes in market share and user churn rate, and determining the market competitiveness score; Obtain the pre-set profit rating weights and market competitiveness rating weights; The pricing strategy score is calculated by weighting the revenue score, the market competitiveness score, the revenue score weight, and the market competitiveness score weight.

6. The vehicle rental operation management method according to claim 3, characterized in that: The process of determining the service strategy score corresponding to each candidate user service strategy includes: For each candidate user service strategy, based on the user's dynamic credit profile, the response time, service accuracy, and complaint handling efficiency are evaluated to determine the service quality score; For each candidate user service strategy, based on the user's dynamic credit profile, predict user repurchase rate, user satisfaction, and user recommendation intention, and determine user retention score.

7. The vehicle rental operation management method according to claim 6, characterized in that: Obtain pre-set service quality score weights and user retention score weights; The service strategy score is calculated by weighting the service quality score, the user retention score, the service quality score weight, and the user retention score weight.

8. The vehicle rental operation management method according to claim 3, characterized in that: The process score for each candidate operation management process is determined based on the scheduling scheme score corresponding to the specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to the specific candidate pricing strategy, and the service strategy score corresponding to the specific candidate user service strategy. Obtain the pre-set scoring weights for scheduling schemes, pricing strategies, and service strategies; The process score is calculated based on the scheduling scheme score corresponding to the specific candidate vehicle scheduling scheme, the pricing strategy score corresponding to the specific candidate pricing strategy, the service strategy score corresponding to the specific candidate user service strategy, the scheduling scheme score weight, the pricing strategy score weight, and the service strategy score weight.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory, wherein the processor, when executing the computer program, implements the method of any one of claims 1-8.

10. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of any one of claims 1-8.

Images