Crowdsourcing service price optimization method and system based on instant travel service

Abstract

The invention provides a crowdsourcing service price optimization method and system based on an instant travel service, and particularly relates to the technical field of data processing, and the scheme comprises the steps: constructing a crowdsourcing distribution model based on the cost and profit of an instant travel service provider; based on the cost of a parcel delivery service provider, constructing a multi-warehouse pickup and delivery model; and based on the initial crowdsourcing service price, performing iterative solution optimization on the crowdsourcing delivery model and the multi-warehouse pickup and delivery model by using a preset algorithm to obtain an optimal crowdsourcing service price. According to the scheme, benefits of a parcel delivery service provider and an instant travel service provider are considered, respective benefit optimization models and solution algorithms are constructed from the perspective of cost and income, and a high-quality solution can be quickly converged, so that the optimal crowdsourcing service price is accurately calculated.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for optimizing the price of crowdsourced services based on on-demand travel services. Background Technology

[0002] Crowd-sourced delivery (Crowd-shipping) is a collaborative model between Parcel Delivery Service Providers (PSPs) and On-demand Mobility Service Providers (OMPs). This model aims to utilize the idle capacity of passenger vehicles to deliver parcels / goods while transporting passengers. Depending on the mode of transport, crowd-sourced delivery models generally take three forms: the first is hiring amateur individuals (such as in-store customers with private cars) as crowdsourced couriers to deliver parcels during their personal journeys; the second is utilizing existing fixed-route public transportation systems (such as bus networks and subway systems) for parcel delivery; and the third is utilizing on-demand mobility service vehicles (such as taxis, ride-hailing services, and future autonomous vehicles) to complete parcel delivery while serving passengers. In this context, designing service prices that meet parcel delivery demand while benefiting both parties is crucial to the success of this new business model. Therefore, determining the optimal crowdsourcing service price to ensure mutual benefit and sustainable cooperation between PSPs and OMPs is of great significance.

[0003] Existing technologies mainly study the crowdsourcing delivery model from the perspective of OMP (Outsourced Management Platform) to determine the crowdsourcing service price. However, this model cannot be separated from the close cooperation between OMP and PSP (Personal Support Service Provider). PSP needs to provide an attractive crowdsourcing service price in order to obtain OMP's delivery service. Therefore, the crowdsourcing service price determined by existing technologies is not accurate enough and cannot well balance the interests of both OMP and PSP. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide a crowdsourcing service price optimization method and system based on on-demand travel services, which aims to solve the problems of insufficient accuracy in crowdsourcing service prices and the inability to properly balance the interests of both OMP and PSP in the prior art.

[0005] To achieve the above objectives, a first aspect of the present invention provides a method for optimizing the price of crowdsourcing services based on on-demand transportation services, comprising:

[0006] Obtain the initial crowdsourcing service price;

[0007] Based on the costs and profits of on-demand transportation service providers, a crowdsourced delivery model is constructed.

[0008] Based on the costs of parcel delivery service providers, construct a multi-warehouse pickup and delivery model;

[0009] Based on the initial crowdsourcing service price, the crowdsourcing delivery model and the multi-warehouse pickup and delivery model are iteratively solved and optimized using a preset algorithm to obtain the optimal crowdsourcing service price.

[0010] Optionally, the construction of a crowdsourced delivery model based on the costs and profits of on-demand transportation service providers includes:

[0011] The total cost of the on-demand mobility service provider is determined based on at least one of the following: pre-acquired fixed operating costs, transportation costs, penalty costs for not meeting passenger travel needs, and passenger compensation costs.

[0012] The total profit of the on-demand transportation service provider is determined based on the pre-obtained profits from delivering packages and transporting passengers.

[0013] Based on the total cost and the total profit, a crowdsourced delivery model is constructed.

[0014] Optionally, the construction of a multi-warehouse pickup and delivery model based on the cost of the parcel delivery service provider includes:

[0015] A multi-warehouse pickup and delivery model is constructed based on at least one of the pre-obtained fixed operating costs, transportation costs, and crowdsourced shipping costs of the package delivery service provider.

[0016] Optionally, the step of iteratively solving and optimizing the crowdsourcing delivery model and the multi-warehouse pickup and delivery model based on the initial crowdsourcing service price to obtain the optimal crowdsourcing service price includes:

[0017] Based on the initial crowdsourcing service price, the crowdsourcing delivery model is solved using a preset first fine-grained tabu search algorithm to calculate the parcel service demand under the initial price;

[0018] Based on the parcel service demand under the initial price, the multi-warehouse pickup and delivery model is solved using a preset second fine-grained tabu search algorithm to calculate the crowdsourcing service price under the initial parcel delivery demand;

[0019] The crowdsourcing service price under the initial package delivery demand is updated using a preset genetic algorithm to obtain an optimized crowdsourcing service price;

[0020] Repeat the steps of calculating the demand for package services and the price of crowdsourcing services, and updating the price of crowdsourcing services, until the preset iteration termination condition is met, and the optimal price of crowdsourcing services is obtained.

[0021] Optionally, before obtaining the optimal crowdsourcing service price, the following steps are also included:

[0022] Obtain passenger travel needs;

[0023] The initial crowdsourcing service price and the passenger's travel demand are used as inputs to the crowdsourcing delivery model, which outputs the optimal service demand and optimal service route for the on-demand travel service provider.

[0024] Optionally, if the optimal service demand cannot meet the total package delivery demand, after outputting the optimal service demand and optimal service route of the on-demand transportation service provider, the method further includes:

[0025] Based on the optimal service demand and the total parcel delivery demand, the remaining parcel delivery demand is determined;

[0026] Based on the remaining parcel delivery demand, the optimal self-operated service strategy of the parcel delivery service provider is determined.

[0027] Optionally, the solution process of the first fine-grained tabu search algorithm and the second fine-grained tabu search algorithm includes:

[0028] Obtain passenger travel demand and total parcel delivery demand;

[0029] The service priority order of the passenger travel demand and the total parcel delivery demand is calculated to obtain the service priority order combination;

[0030] Based on all combinations of the service sequence, the delivery profit of the package delivery service provider is calculated using the first fine-grained tabu search algorithm, and the delivery cost of the package delivery service provider is calculated using the second fine-grained tabu search algorithm.

[0031] The crowdsourcing service price is determined based on the delivery profit, the delivery cost, and a preset fine-grained threshold.

[0032] A second aspect of the present invention provides a crowdsourcing service price optimization system based on on-demand travel services, the system comprising:

[0033] The information acquisition module is used to obtain the initial crowdsourcing service price;

[0034] The crowdsourced delivery model building module is used to build crowdsourced delivery models based on the costs and profits of on-demand transportation service providers;

[0035] A multi-warehouse pickup and delivery model building module is used to build multi-warehouse pickup and delivery models based on the costs of parcel delivery service providers;

[0036] The crowdsourcing service price decision module is used to iteratively solve and optimize the crowdsourcing delivery model and the multi-warehouse pickup and delivery model based on the initial crowdsourcing service price, and obtain the optimal crowdsourcing service price.

[0037] A third aspect of the present invention provides a smart terminal, the smart terminal including a memory for storing executable instructions; and a processor for calling and running the executable instructions in the memory to execute any of the steps of the above-described crowdsourcing service price optimization method based on instant travel services.

[0038] A fourth aspect of the present invention provides a computer-readable storage medium storing program instructions that, when executed by a processor, implement any of the steps of the above-described crowdsourcing service price optimization method based on on-demand travel services.

[0039] Compared with existing technologies, the beneficial effects of this solution are as follows:

[0040] This invention constructs a crowdsourced delivery model based on the costs and profits of on-demand transportation service providers; it also constructs a multi-warehouse pickup and delivery model based on the costs of parcel delivery service providers; and iteratively optimizes both the crowdsourced delivery model and the multi-warehouse pickup and delivery model using a pre-defined algorithm based on an initial crowdsourced service price, thereby obtaining the optimal crowdsourced service price. This scheme considers the interests of both parcel delivery service providers and on-demand transportation service providers, constructing their respective benefit optimization models from the perspectives of cost and revenue, and using a fine-grained tabu search algorithm to solve the models. This allows for rapid convergence to high-quality solutions, thereby accurately calculating the optimal crowdsourced service price. It effectively balances the interests of both OMP (On-Demand Platform) and PSP (On-Demand Service Provider), providing important guidance for the formulation and sustainable operation of crowdsourced delivery models for on-demand transportation services. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0042] Figure 1 Here is a simplified flowchart of the crowdsourcing service price optimization method based on on-demand travel services of the present invention;

[0043] Figure 2This is a flowchart of the crowdsourcing service price optimization method based on on-demand travel services of the present invention.

[0044] Figure 3 This is a schematic diagram of the crowdsourcing service price optimization system module based on on-demand travel services of the present invention;

[0045] Figure 4 This is a schematic diagram of the intelligent terminal structure of the present invention. Detailed Implementation

[0046] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0047] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0048] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0049] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] 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.

[0052] Existing technologies primarily study the decision-making process for crowdsourced delivery models from the perspective of Outsourcing Platforms (OMPs). However, this model relies heavily on close cooperation between OMPs and Service Providers (SPs), requiring PSPs to offer attractive crowdsourced service prices to secure OMP delivery services. Therefore, the crowdsourced service prices determined by existing technologies do not adequately balance the interests of both OMPs and PSPs. This invention aims to establish a close cooperative relationship between OMPs and PSPs, considering that PSPs need to offer attractive crowdsourced service prices to obtain OMP services. OMPs then decide which package delivery needs can be served, and unserved package delivery needs must be handled by the PSP's own delivery fleet. In this scenario, the PSP needs to determine the crowdsourced service price while considering OMP's decisions. The game between PSPs and OMPs forms a Stackelberg game, in which PSPs are dominant in determining package delivery prices and whether to outsource package delivery needs to OMPs. Essentially, the PSP acts as the leader, seeking the optimal crowdsourced service price, while OMP acts as the follower, determining the optimal package delivery needs to be served at a given crowdsourced delivery price.

[0053] Based on this, the technical problem to be solved by this invention is to determine the optimal crowdsourcing service price for the PSP and the service routes for both the PSP and OMP, considering the game between the PSP and OMP, so that the PSP can complete all package delivery needs in the most cost-effective way, while the OMP can obtain the maximum total profit from the passengers served and the crowdsourcing delivery needs. This problem is formulated as a bilevel programming model, including constructing a multi-warehouse pickup and delivery model from the PSP's perspective and a crowdsourcing delivery model from the OMP's perspective, and using a fine-grained tabu search algorithm and a genetic algorithm to solve the model and obtain the optimal crowdsourcing service price. Based on this bilevel programming model, embodiments of this invention provide a crowdsourcing service price optimization method based on on-demand transportation services. This method is deployed on electronic devices such as computers and servers and applied to crowdsourcing delivery scenarios requiring on-demand transportation services. It addresses the situation where close cooperation between the OMP and PSP is considered, allowing the PSP to provide attractive crowdsourcing service prices while enabling the OMP to obtain better delivery services. The crowdsourcing package delivery needs of this invention only involve packages within cities that are easily transportable using passenger vehicles. Figure 1 and Figure 2 As shown, the steps of the method in this embodiment include:

[0054] Step S100: Obtain the initial crowdsourcing service price;

[0055] Specifically, an initial crowdsourcing service price is obtained. This initial crowdsourcing service price can be determined using existing technology or based on the experience of professionals. This reduces the need for subsequent iterative optimization of the crowdsourcing service price and speeds up processing efficiency.

[0056] Step S200: Based on the costs and profits of on-demand transportation service providers, construct a crowdsourced delivery model;

[0057] To fulfill package delivery needs at minimal cost, the PSP will collaborate with the OMP, outsourcing package delivery tasks to the OMP. A certain monetary reward will be paid per unit of package load. In practice, the price for package delivery is determined based on the required package load; therefore, this embodiment defines a load-based crowdsourcing service price, u. The OMP will use on-demand vehicles to decide which package delivery needs share transportation services with passenger ride requests. The collaboration model between the PSP and OMP involves the PSP providing the OMP with package delivery needs and crowdsourcing service prices. The OMP then considers both crowdsourcing and passenger needs to determine the optimal service for each type of package and passenger to maximize its total profit. Furthermore, any unserved crowdsourcing needs and their corresponding costs are fed back to the PSP. Ultimately, the PSP makes its own delivery service decisions to meet these unserved crowdsourcing needs. As can be seen, the game between PSP and OMP forms a Steinberg game, where PSP, as the leader, considers OMP's optimal decision and aims to propose the optimal crowdsourcing service price, while OMP, as the follower, aims to make the optimal decision, i.e., determine the acceptable package delivery demand at a given price. By cooperating with OMP, PSP can meet crowdsourcing delivery demand through OMP's crowdsourcing delivery service or its own service.

[0058] Based on the above analysis, this embodiment constructs a crowdsourcing delivery model based on the costs and profits of on-demand transportation service providers. Specifically, it calculates the total cost by combining one or more of the on-demand transportation service provider's fixed operating costs, transportation costs, penalty costs for not meeting passenger travel needs, and passenger compensation costs, or even by combining one or more of these costs with other possible costs. The total profit is then calculated by combining the on-demand transportation service provider's profit from delivering packages and its profit from transporting passengers. The profit from transporting passengers represents the passenger fare minus the compensation given to passengers for delivering packages. Finally, based on the on-demand transportation service provider's total costs and total profits, a crowdsourcing delivery model is constructed. This model represents the difference between total profit and total cost, reflecting the net profit that the on-demand transportation service provider can obtain through crowdsourcing delivery services.

[0059] Step S300: Based on the costs of parcel delivery service providers, construct a multi-warehouse pickup and delivery model;

[0060] Similarly, based on the costs of parcel delivery service providers, a multi-warehouse pickup and delivery model is constructed. This involves using one or more of the fixed operating costs, transportation costs, and crowdsourcing costs of the parcel delivery service provider, or even combining one or more of these costs with other possible costs, to construct a multi-warehouse pickup and delivery model. This multi-warehouse pickup and delivery model is used to represent the total cost paid by the parcel delivery service provider for delivering parcels.

[0061] Step S400: Based on the initial crowdsourcing service price, use a preset algorithm to iteratively solve and optimize the crowdsourcing delivery model and the multi-warehouse pickup and delivery model to obtain the optimal crowdsourcing service price.

[0062] Granular Tabu Search Algorithms (GTSs) are a variant of the tabu search algorithm. They improve search efficiency and solution quality by using a fine-grained neighborhood structure. A fine-grained neighborhood refers to considering only moves or changes that are likely to produce high-quality solutions during the search process, thereby reducing the size of the search space and improving search efficiency. A fine-grained neighborhood can be seen as an efficient implementation of the candidate list strategy proposed by the tabu search algorithm. In each iteration, the GTS only evaluates potentially promising moves to reduce the possibility of short-sighted searches. The structure of the multi-granularity neighborhood can dynamically change during the algorithm's evolution to diversify the search. This dynamism helps the algorithm maintain flexibility during the search process, better adapt to changes in the problem, and enable the algorithm to converge to a high-quality solution more quickly. Therefore, this embodiment uses the GTS to iteratively solve and optimize the crowdsourcing delivery model and the multi-warehouse pickup and delivery model to obtain the optimal crowdsourcing service price. Specifically, the search space and candidate list are initialized using the initial crowdsourcing service price. Candidate solutions are generated by making limited and promising moves based on the fine-grained neighborhood of the current solution. The crowdsourcing service price is updated by comparing it with the current crowdsourcing service price and the historical best crowdsourcing service price. The above steps are repeated until the termination condition is met (such as reaching the maximum number of iterations or finding a crowdsourcing service price that satisfies both PSP and OMP), and the optimal crowdsourcing service price is obtained.

[0063] In this embodiment, by considering the interests of both the parcel delivery service provider PSP and the on-demand transportation service provider OMP, a crowdsourcing delivery model suitable for on-demand transportation service providers and a multi-warehouse pickup and delivery model suitable for parcel delivery service providers are constructed from the perspectives of cost and benefit, respectively. The model is solved using a fine-grained tabu search algorithm, which can quickly converge to a high-quality solution, thereby accurately calculating the optimal crowdsourcing service price and demonstrating excellent game theory results.

[0064] In one embodiment, assume a PSP (Package Service Provider) provides daily parcel delivery services within a city area using its own fleet of trucks. This city area has several warehouses (represented by set H) for parking the trucks (represented by set K). For each delivery request received by the PSP, there are corresponding pick-up locations, time windows, loading service durations, drop-off locations, time windows, unloading service durations, and parcel volumes, etc., to fulfill the delivery request through the settings of these nodes. Within the same city area, assume an OMP (On-Demand Platform) provides passenger transport services using several on-demand vehicles (represented by set V), where several stations (represented by set S) in set V are scattered throughout the area to park these vehicles. The OMP receives several passenger ride requests, each with an origin, a destination, a pick-up time window at the origin, a drop-off time window at the destination, and a number of passengers.

[0065] Let Ω f,o and Ω f,d Ω represents the set of pickup and drop-off locations for package delivery requests, respectively. p,o and Ω p,d Let f represent the set of origins and destinations for passenger travel requests, respectively, where f represents a package, p represents a passenger, o represents the pickup / departure point, and d represents the drop-off / drop-off point. For ease of representation, let m and n represent the quantity of crowdsourced delivery requests and passenger requests, respectively.

[0066] According to Ω f,o ={1,2,L,m},Ω p,o ={m+1,m+2,L,σ},Ω f,d ={σ+1,σ+2,L,σ+m}, Ω p,d The ordered arrangement of {σ+m+1,σ+m+2,L,2σ}, where σ=m+n. Define Ω=Ω f,o ∪Ω p,o ∪Ω f,d ∪Ω p,d Therefore, the drop-off point / destination of a package / passenger can be represented by the pickup point / origin. For simplicity, we use the index i∈Ω of the pickup point / origin of the package / passenger's request. f,o UΩ p,oTo represent the corresponding package / passenger demand, let [e i ,l i ] indicates the demand i∈Ω f,o UΩ p,o The pickup time window, of which e i and l i These represent the earliest and latest pickup times, respectively. Correspondingly, [e i+σ ,l i+σ ] represents package / passenger demand i∈Ω f,o UΩ p,o Drop-off time window; and Let i ∈ Ω represent the crowdsourced delivery demand respectively. f,o Parcel volume and passenger demand i∈Ω p,o The number of people; d i and d i+σ They are respectively represented as packages i∈Ω f,o Service time required for loading and unloading packages.

[0067] For OMP, each vehicle v∈V initially parks in a parking space with a capacity of Q. s For any station s∈S, after the service ends, it can return to any station s∈S. The maximum usage time of vehicle v∈V is expressed as... Each vehicle, v∈V, can simultaneously serve multiple passenger and crowdsourced delivery needs. The carrying capacity of passengers and packages is respectively represented by... and This indicates that, to ensure service quality for passengers, the longest possible travel time will be considered. In addition, each passenger i∈Ω p,o The maximum total number of stops allowed between the origin and destination to accommodate additional passengers or packages is ξ. max Furthermore, if detours are necessary, passengers must be compensated with a fee c based on the detour time. Let η represent the fixed operating cost per vehicle, and R... i Indicates the passenger demand for services i∈Ω p,o The gains obtained, P i The penalty for refusing a passenger's request is represented by t, where t represents the travel time and cost of an on-demand vehicle from location i to j. i,j and c i,j The goal of OMP is to determine the optimal package and passenger service demand, and the corresponding service routes, at the price of the crowdsourced service offered, in order to: (i) meet the service time window, the longest travel time, and the maximum number of stops (if applicable); (ii) respect parking lot capacity, vehicle capacity, and maximum service time; and (iii) maximize total profit.

[0068] For PSP, unserved packages will be handled by a fleet of self-operated trucks K, where each truck k∈K is initially parked at a parking capacity of Q. h The service can be completed at any parking lot h∈H, and the truck can return to any parking lot h∈H. The fixed operating cost, parcel carrying capacity, and maximum service time of the truck k∈K are represented by ψ, Q, and Q, respectively. k and make and κ i,j Let $\mathbf$ and $\mathbf$ represent the travel time and cost of the vehicle from location $i$ to location $j$. The goal of PSP is to find the lower limit of the price range. u and price ceiling The crowdsourcing service price and self-service routes are determined in order to: (i) meet the service time window for delivery demand; (ii) ensure parking lot capacity, truck capacity and maximum service time; and (iii) minimize the total cost, including crowdsourcing costs and self-operated costs.

[0069] The OSP problem studied in this embodiment considers the game between PSP and OMP, determining the optimal crowdsourcing service price and the PSP's self-operated service routes and the OMP's crowdsourcing service routes. It can be seen that the OMP's decisions are influenced by the PSP's pricing decisions, and the OMP's decisions directly affect the availability of services to the PSP, thus affecting the OMP's pricing decisions. In other words, the OSP problem studied in this embodiment has the characteristic of mutual influence between the cooperating parties, that is, a bridge is built between the PSP and OMP through pricing decisions. In this case, describing this game using a single optimization model may be challenging. Therefore, this embodiment considers the OSP problem within a two-layer framework based on the game between PSP and OMP, that is, it is established as a two-layer programming model. The upper-layer programming model solves the multi-depot pickup and delivery problem with pricing for PSP (MPDP-P), and the lower-layer programming model solves the crowd-shipping problem with ridesharing for OMP (CSP-R).

[0070] The following examples illustrate in detail the construction process of the upper-level and lower-level planning models:

[0071] In a preferred embodiment, a crowdsourced delivery model, namely the lower-level CSP-R model, is constructed based on the costs and profits of on-demand transportation service providers.

[0072] Based on at least one of the pre-acquired fixed operating costs, transportation costs, penalty costs for not meeting passenger travel needs, and passenger compensation costs of the on-demand transportation service provider, determine the total cost of the on-demand transportation service provider; based on the pre-acquired profits of the on-demand transportation service provider in delivering packages and in transporting passengers, determine the total profit of the on-demand transportation service provider; based on the total cost and the total profit, construct a crowdsourced delivery model.

[0073] Specifically, given the crowdsourcing service price u and the crowdsourcing delivery demand Ω f,o and passenger demand Ω p,o In this context, a crowdsourced delivery model is constructed, combining u and Ω. f,o and Ω p,o As input to the crowdsourced delivery model, the output is the optimal service demand and optimal service route for the on-demand transportation service provider, maximizing the total profit obtained by OMP from service packages and passenger demand. Since multiple parking locations are considered in the lower-level problem, for ease of modeling, the site set S is replicated, and the original site set is named S0. o The copied site set is named S d S o and S d Let S represent the set of origin stations and the set of destination stations for the vehicles, respectively. In fact, S... o and S d It consists of the same sites, duplicate sites S d The parking capacity is equal to the corresponding original site S o The parking capacity. Define the lower-level CSP-R model on a fully directed graph, i.e.:

[0074] G1 = (N1, A1),

[0075] N1=ΩU S o US d ,

[0076] A1=(S o ×(Ω p,o UΩ f,o ))U(Ω×Ω)U((Ω p,d UΩ f,d )×S d ).

[0077] Each node i∈N1 in the network has a package load. Passenger numbers Service time window [e i ,l i Service duration d i Income R i and punishment P iFor each arc (i,j)∈A1, the travel time from node i to j using an on-demand vehicle is t. i,j And travel costs are c i,j , The information is as follows:

[0078] Parcel volume / passenger volume:

[0079] Service Hours:

[0080] income:

[0081] Penalty:

[0082] In addition, the following variables are further defined:

[0083] z i : A binary decision variable, if the service demand i∈Ω p,o UΩ f,o Then z i It equals 1, otherwise it equals 0;

[0084] For a binary decision variable, if vehicle v∈V travels directly from node i to node j... but It is 1 if it is true, otherwise it is 0;

[0085] A continuous variable representing the time when vehicle v starts serving at node i∈N1;

[0086] A continuous variable, representing passengers i∈Ω in vehicle v∈V. p,o Total travel time;

[0087] A continuous variable representing the number of passengers in vehicle v∈V after the service ends at node i∈N1;

[0088] A continuous variable representing the load packaged in vehicle v∈V after node i∈N1 provides service;

[0089] An integer variable representing the order of node i∈N1 in the vehicle v service sequence.

[0090] Based on the above description, the model of the lower-level CSP-R can be expressed as follows:

[0091]

[0092] The following constraints must be met:

[0093]

[0094]

[0095]

[0096] In formula (1), the objective function represents maximizing the total profit of OMP, which is the difference between demand revenue and total cost. The total cost includes fixed operating costs, transportation costs, penalty costs for unmet passenger demand, and passenger compensation costs. Constraints (2) and (3) are inbound and outbound flow balance constraints. Constraint (2) specifies that each vehicle departs from a certain station and eventually returns to any station. Constraint (3) represents the inbound and outbound flow balance constraints for any node other than the vehicle's origin and destination. Constraint (4) ensures that each package delivery demand or passenger ride demand can be served at most once. Constraint (5) ensures that the pick-up and drop-off operations for each package or passenger demand should be served by the same vehicle. Constraint (6) updates the start time of each demand on the vehicle route. Constraint (7) specifies the order of pickup and delivery operations for each demand. Constraint (8) specifies the service time window for service demands. Constraint (9) calculates the actual ride time for each passenger. Constraint (10) specifies the maximum travel time that passengers can tolerate; constraint (11) limits the maximum usage time of each vehicle; constraints (12) and (13) update the number of passengers on board and the load of parcels in the vehicle, respectively; constraints (14) and (15) limit the capacity to carry passengers and parcels, respectively; constraints (16) and (17) are the capacity constraints of the parking lot; constraint (16) specifies that the total number of initial dispatched vehicles departing from the parking lot should not exceed the maximum parking capacity of the parking lot, and constraint (17) ensures that the number of vehicles ultimately parked in the parking lot does not exceed the maximum parking capacity of the parking lot; constraints (18) and (19) are used to calculate the travel sequence of demand, where M can be set to 2(m+n+1); constraint (20) specifies the maximum number of stops between the departure point and the destination of passenger demand; constraints (21)-(24) define the feasible domain of the corresponding variables.

[0097] This embodiment establishes a lower-level CSP-R model by comprehensively considering the difference between the demand revenue and total cost of the on-demand mobility service provider OMP, and determines the optimal service demand and optimal service route, thereby ensuring that OMP can maximize the total profit obtained from service packages and passenger demand.

[0098] In a preferred embodiment, a multi-warehouse pickup and delivery model, i.e., an upper-level MPDP-P model, is constructed based on the costs of the parcel delivery service provider, including:

[0099] Based on the optimal service demand decision of the on-demand service provider OMP obtained from the lower-level CSP-R model, if the optimal service demand decision cannot meet the total parcel delivery demand, then the unserved crowdsourced delivery demand under the crowdsourced service price is calculated based on the total parcel delivery demand and the delivery service demand. Then, based on the unserved crowdsourced delivery demand, the fixed operating costs, transportation costs, and crowdsourced shipping costs of the parcel delivery service provider, and a multi-warehouse pickup and delivery model, the remaining parcel delivery demand is determined based on the optimal service demand and the total parcel delivery demand. Based on the remaining parcel delivery demand, the optimal self-operated service strategy of the parcel delivery service provider is determined, including the optimal crowdsourced service price and self-operated service route, to minimize the total cost of the PSP, i.e., to minimize the cost of the parcel delivery service provider PSP.

[0100] Specifically, let and Let f represent the pickup location and drop-off location for unprocessed crowdsourced delivery requests, respectively. Here, f represents the package, o represents the pickup location, and d represents the drop-off location. * indicates the optimal solution obtained by solving the lower-level problem, and rej indicates rejection. Specifically, and This represents the pickup and drop-off locations for unprocessed (i.e., rejected) crowdsourced delivery requests obtained by solving the lower-level CSP-R problem. Similar to the lower-level problem, for ease of modeling, the warehouse set H is copied to Hwarehouse-set-H. o and H d H o and H d H represents the set of originating and destination warehouses for the trucks. o and H d Including the same warehouse. Then define the upper-level MPDP-P model on the fully directed graph, i.e.:

[0101] G2 = (N2, A2),

[0102]

[0103] Each node i∈N2 in the network has a wrapper payload. Service time window [e i ,l i ] and service duration d i For each arc and truck traveling from node i to j, (i,j)∈A2, the travel time and travel cost are... and κ i,j , It contains the following information: Package load Service duration d i =0,

[0104] In addition, the following variables are further defined:

[0105] For a binary decision variable, if the truck k∈K travels directly from node i to node j... The value is 1 if it is 1, otherwise it is 0.

[0106] A continuous variable representing the time when vehicle k∈K starts service at node i∈N2;

[0107] A continuous variable representing the load of packages inside the truck after node k∈K serves i∈N2.

[0108] Based on the above description, the upper-level MPDP-P model can be represented as:

[0109]

[0110] The following constraints must be met:

[0111]

[0112]

[0113] In formula (25), the objective function is to minimize the total cost of the PSP, including the fixed operating cost of the trucks, transportation costs, and crowdsourcing costs paid to the OMP, which are influenced by the results of the CSP-R model. Constraints (26) and (27) represent inbound and outbound flow balance constraints. Constraint (26) specifies that each truck departs from a vehicle segment and eventually returns to any vehicle segment after completing its delivery task, while constraint (27) ensures flow balance at any intermediate node. Constraints (28) and (29) ensure that each package delivery request should be served exactly once by one truck. It is worth noting that a vehicle may revisit the same physical location, albeit with a different index, as it belongs to the same location where it recently stopped. Constraint (30) updates the start time of service for each crowdsourced delivery request on the truck route. Constraint (31) ensures the order of pickup and delivery operations for each package delivery request. Constraints (32) and (33) specify the service time window for the service request and the maximum service time for each vehicle, respectively. Constraint (34) updates the vehicle load. Constraint (35) limits the capacity of the trucks. Constraints (36) and (37) specify the warehouse capacity. It is worth noting that each truck can reach any parking lot (including the one from which it departs) unless the parking capacity is not exceeded. Constraint (38) limits the upper and lower limits of the crowdsourcing service price of the PSP. Constraints (39)-(40) define the feasible domain of the corresponding decision variables.

[0114] Based on the results of the CSP-R model, this embodiment constructs a multi-warehouse pickup and delivery model by considering the cost of package delivery service providers (PSPs) from multiple perspectives, and determines the optimal crowdsourcing service price and self-operated service route, which can minimize the total cost of PSPs.

[0115] In a preferred embodiment, based on the initial crowdsourcing service price, the crowdsourcing delivery model and the multi-warehouse pickup and delivery model are iteratively solved and optimized using a preset algorithm to obtain the optimal crowdsourcing service price, including:

[0116] Based on the initial crowdsourcing service price, a preset first fine-grained tabu search algorithm is used to solve the crowdsourcing delivery model and calculate the package service demand under the initial price. Based on the package service demand under the initial price, a preset second fine-grained tabu search algorithm is used to solve the multi-warehouse pickup and delivery model and calculate the crowdsourcing service price under the initial package delivery demand. A preset genetic algorithm is used to update the crowdsourcing service price under the initial package delivery demand to obtain an optimized crowdsourcing service price. The steps of calculating the package service demand and crowdsourcing service price and updating the crowdsourcing service price are repeated until a preset iteration termination condition is reached to obtain the optimal crowdsourcing service price.

[0117] Specifically, the OSP problem considered in this embodiment is a two-layer optimization problem, with the lower-layer CSP-R model nested within the upper-layer MPDP-P model. The upper-layer MPDP-P model is an extension of the multi-warehouse vehicle routing problem, while the lower-layer CSP-R model is a variant of the DARP (Dial-a-ride problem). Both problems have been proven to be NP-hard, making the OSP problem unsolvable by commercial solvers. However, it can be seen from the two-layer framework of the OSP problem that once the crowdsourcing service price is determined, the upper-layer problem simplifies to MPDP, i.e., determining the PSP's self-operated service routes to meet the unserved crowdsourcing delivery demand obtained by solving the lower-layer CSP-R at the aforementioned price. Based on this, this embodiment proposes a customized Interactive Hybrid (IH) algorithm. Specifically, firstly, a profit-oriented granular tabu search algorithm (TS-P) proposed in this embodiment is used to solve the lower-level CSP-R model (i.e., the crowdsourcing delivery model) to obtain the unserved crowdsourcing delivery demand at the initial price. Then, a cost-oriented granular tabu search algorithm (TS-C) proposed in this embodiment is used to solve the simplified MPDP-P model (i.e., the simplified multi-warehouse pickup and delivery model) at the initial price. Finally, a preset genetic algorithm (GA) is used to update the crowdsourcing service price. Finally, the process of solving the lower-level CSP-R model, solving the simplified MPDP-P model, and updating the crowdsourcing service price is iteratively executed until the maximum number of iterations is reached or a crowdsourcing service price agreed upon by both parties is obtained, thus obtaining the optimal crowdsourcing service price and corresponding service route for PSP and OMP. This embodiment does not specifically limit the type of genetic algorithm used; any algorithm capable of updating the crowdsourcing service price is acceptable. The overall framework of the IH algorithm is as follows: Figure 2 As shown.

[0118] In this embodiment, the IH algorithm combines two fine-grained tabu search algorithms and a genetic algorithm. It fully leverages the advantages of fine-grained tabu search algorithms in solving pickup and delivery problems, namely their search efficiency and ability to guarantee solution quality. Simultaneously, since the genetic algorithm is a population-based metaheuristic algorithm, it simulates biological genetic operations and retains superior individuals through fitness values ​​to determine the total cost of the PSP from the OSP problem. This allows for the selection of the optimal crowdsourcing service price by generating multiple crowdsourcing service prices. Therefore, the combination of GTS and GA can ensure that the total cost of the PSP is minimized while guaranteeing that the OMP obtains the optimal crowdsourcing service price, while fully considering the interaction between the PSP and OMP.

[0119] In a preferred embodiment, the solution process of the first fine-grained tabu search algorithm and the second fine-grained tabu search algorithm includes:

[0120] Obtain passenger travel demand and total parcel delivery demand;

[0121] The service priority order of the passenger travel demand and the total parcel delivery demand is calculated to obtain the service priority order combination;

[0122] Based on all combinations of the service sequence, the delivery profit of the package delivery service provider is calculated using the first fine-grained tabu search algorithm, and the delivery cost of the package delivery service provider is calculated using the second fine-grained tabu search algorithm.

[0123] The crowdsourcing service price is determined based on the delivery profit, the delivery cost, and a preset fine-grained threshold.

[0124] Specifically, since the lower-level CSP-R model is used to find service routes that maximize profits, this embodiment proposes a profit-oriented first fine-grained tabu search algorithm to explore promising mobility operations that can generate high-profit solutions, thereby obtaining a profit-oriented fine-grained strategy. Therefore, this embodiment will consider the profit of vehicles serving demands i and j. Defined as the sequence Π starting from demand i. Considering the pickup and delivery operations for each demand, there are three possible sequences: Π1=(i,i+σ,j,j+σ), Π2=(i,j,i+σ,j+σ) and Π3=(i,j,j+σ,i+σ), ensuring that demand i is served before j.

[0125] make Let represent the average profit of vehicles with service demand i up to the j-th sequence in the three sequences. The larger the value of , the more profitable it is to use a vehicle to satisfy demand j after demand i. Therefore, a promising mobility operation is defined as one that enables a vehicle to serve demand i before j. P gran This represents a fine-grained threshold for profit. For the lower-level CSP-R model, demand i and j can be different types of demand, such as package delivery demand or passenger travel demand. Therefore, calculations are performed based on the following four cases.

[0126] Scenario 1 (Both demand i and j are passenger travel demands):

[0127]

[0128] Scenario 2 (Both demand i and j are parcel delivery demands):

[0129]

[0130] Scenario 3 (Demand i is passenger travel demand, while demand j is parcel delivery demand):

[0131]

[0132] Scenario 4 (Demand i is parcel delivery demand, and demand j is passenger travel demand):

[0133]

[0134] Among them, c Π This represents the total transportation cost generated along the sequence Π. For example, for a sequence Π1=(i,i+σ,j,j+σ), we have c Π =c i,i+σ +c i+σ,j +c j,j+σ If sequence Π violates any relevant constraints, namely the demand time window constraint (Equation (8)), the passenger riding time constraint (Equation (10)), the vehicle working time constraint (Equation (11)), or the vehicle load constraint (Equation (10)), let... That is, demand i and demand j cannot be served by the same vehicle in the corresponding order. The formula for calculating average profit is:

[0135]

[0136] Where ζ Π This is an auxiliary variable; it is 1 if it is true, and 0 otherwise. set up This means that demands i and j cannot be served by the same vehicle starting from demand i. Through calculation... And compare with a given profit fine-grained threshold P gran The promising move operations can be calculated as follows: Then, a neighborhood solution is generated on the current solution through a move operation, which also meets the move operation requirements, i.e., service requirement i is served before j.

[0137] Compared to traditional methods that search all possibilities of the selected demand and insertion position, fine-grained strategies can obtain reduced neighborhood solutions by searching only promising move operations. gran A larger value means fewer movement operations will be explored, reducing time costs and resulting in a more limited search solution. In the fine-grained strategy exploration process, the profit fine-grained threshold P... gran As the fine-grained threshold P is dynamically adjusted iteratively, gran Initially set to Then gradually decrease P gran ←P gran -δ, until the lower bound P is reached. gran The lower bound P gran According to the definition For example Where χ p This represents the parameters that control the movement operation.

[0138] Since the simplified MPDP-P model at the upper level is used to find the lowest-cost service routes to meet all unserved package delivery needs, this embodiment proposes a cost-oriented, fine-grained strategy to explore promising actions to generate low-cost solutions. Specifically, this embodiment defines... To serve vehicles with demands i and j, the order is Π, starting from demand i. The cost. Similarly, there are three possible sequences that serve the i-th demand before the j-th demand. Let This represents the average cost of the vehicles serving the i-th demand before the j-th demand in the three sequences. The smaller the value, the more cost-effective it is to use a vehicle to complete demand j after demand i. Therefore, A promising movement operation is defined as a movement operation that allows a vehicle to serve a demand i before vehicle j, where C gran This represents a fine-grained threshold for cost. For the simplified MPDP-P model, It can be represented as:

[0139]

[0140] Among them κ Π This represents the total transportation cost of the truck along the sequence Π. For example, for a sequence Π1=(i,i+σ,j,j+σ), κ Π =κi,i+σ +κ i+σ,j +κ j,j+σ Similarly, if sequence Π violates any relevant constraints, namely the time window constraint for setting demand (Equation (33)), the working time constraint for trucks (Equation (34)), or the load capacity constraint for trucks (Equation (36)), This means that demands i and j cannot be served by the same vehicle in the corresponding order. The method for calculating average cost is as follows:

[0141]

[0142] Among them ι Π It is an auxiliary variable, if Otherwise Π =0. If set up This means that demands i and j cannot be served by the same vehicle starting from demand i. Through calculation... and a given cost fine-grained threshold C gran It is possible to calculate promising movement operations. Then, based on the current solution, a move operation is performed to generate a neighborhood solution while satisfying the move operation requirements. This is related to dynamically setting P. gran Similarly, the initial cost is set to a fine-grained threshold C. gran for And gradually increase C gran ←C gran +δ, until the upper bound is reached. The upper bound According to the definition For example Where χ c This represents the parameters that control the movement operation.

[0143] In summary, the method of this invention differs from existing methods that only consider operational-level decision-making problems, such as route optimization. By taking into account the game between PSP and OMP, it proposes a new benefit optimization model and solution algorithm to solve for the optimal crowdsourcing service price and the service routes of PSP and OMP. This provides a better crowdsourcing service price for the operation of the crowdsourcing delivery model and can play an important guiding role in the formulation and sustainable operation of the crowdsourcing delivery model for on-demand travel services.

[0144] The effectiveness of the proposed model and algorithm is evaluated through a computational experiment. The experiment uses the Solomon dataset, a pre-existing resource for time-window vehicle routing problems. Each instance in the dataset contains a warehouse and multiple delivery requests, along with detailed information related to delivery location, demand load, service time window, and service duration. To ensure that each package or passenger request is associated with its corresponding pickup and drop-off requests, data is selected from the C1_2_1 case in the Solomon dataset and further adjusted. Specifically, m and n requests are selected from dataset C1_2_1 as package and passenger pickup requests, respectively, and the same number of requests with different locations are selected from the same set as corresponding drop-off requests. To ensure the validity of the requests, parameters such as service time window, load, number of passengers, and service duration are introduced. All requests within set C1_2_1 operate within a time window of 0–1351, which serves as the operation period. For package pickup requests, the original service time window and service duration of 90 are maintained. Each package pickup request is assigned a positive integer package load, randomly generated within the range [1, 5]. For pick-up requests, the original service time window is retained, and a positive integer number of passengers is assigned to each request, randomly generated within the range [1,4], with a service duration of 0. For return requests, the earliest service time is calculated as the sum of the earliest pick-up time and the travel time from the pick-up location to the return location. Similarly, the latest service time is determined by adding the longest delivery time or travel time to the latest pick-up time. The package load and the number of passengers for drop-off requests are set to the negative values ​​of their corresponding pick-up requests. The service duration for package delivery requests is set to 90, while the service duration for passenger delivery requests is zero. In addition, 5 locations are randomly selected from set C1_2_1 as warehouses for parking trucks, and 20 trucks are randomly assigned to these warehouses, ensuring that each warehouse holds no more than 5 trucks; 5 locations are randomly selected from set C1_2_1 as parking spots for on-demand vehicles, and 12 vehicles are randomly distributed among these parking spots, with each parking spot accommodating no more than 5 vehicles. Detailed parameter settings for the TS algorithm and IH algorithm under different instances are shown in Table 1. The maximum number of iterations for both the upper and lower level problems is set to 200, and the population is set to 20 generations. Each generation consists of 10 crowdsourcing service prices used to update the prices. Based on experimental testing, the penalty coefficient for violating the constraints is set as follows: ω = (Time window constraint for violating the demand) tw =100, Violation of working hour constraints for trucks and on-demand vehicles ω wtk =ω wtv =20. Violation of truck load capacity constraints ω cap_k =30. Violation of passenger travel duration constraints ω rd=20. Violation of vehicle load capacity constraints Penalty coefficient for violating the time window constraints of on-demand mobility vehicles It should be noted that since there are no explicit unit restrictions for time data in the Solomon dataset, the unit of time data can be set to minutes, seconds, etc., according to actual needs. This example tests the algorithm performance by setting the unit of time data to minutes.

[0145] Then, the performance of the proposed IH algorithm is evaluated using different combinations of package and passenger demand. The results of the classic tabu search method (TS algorithm) and the IH algorithm with the fine-grained strategy proposed in this invention are compared to obtain the optimal crowdsourcing service price (u) for the total number of passenger demand (n) and the total number of package delivery demand (m) under different conditions. * The results for the total computation time (CPU) are shown in Table 1.

[0146] Table 1:

[0147]

[0148] As shown in Table 1, with the same number of iterations for all instances, the average computation time of the TS algorithm is 62.2 minutes, while the average computation time of the method proposed in this invention is 25.8 minutes. It can be seen that the method proposed in this invention calculates the optimal crowdsourcing service price in a shorter computation time. In other words, in the same numerical experiment, the TS algorithm requires a lot of time to explore all possible movement operations. In contrast, the method proposed in this invention adopts a fine-grained strategy to find high-quality solutions in a shorter time, effectively combining cost-oriented and profit-oriented fine-grained approaches, and demonstrating high efficiency and reliability in terms of computation time and optimization efficiency.

[0149] like Figure 3 As shown, corresponding to the above-mentioned crowdsourcing service price optimization method based on on-demand travel services, this embodiment of the invention also provides a crowdsourcing service price optimization system based on on-demand travel services, the above-mentioned crowdsourcing service price optimization system based on on-demand travel services includes:

[0150] Information acquisition module 310 is used to obtain the initial crowdsourcing service price;

[0151] Crowdsourced delivery model building module 320 is used to build crowdsourced delivery models based on the costs and profits of on-demand transportation service providers;

[0152] Multi-warehouse pickup and delivery model building module 330 is used to build multi-warehouse pickup and delivery models based on the costs of parcel delivery service providers;

[0153] The crowdsourcing service price decision module 340 is used to iteratively solve and optimize the crowdsourcing delivery model and the multi-warehouse pickup and delivery model based on the initial crowdsourcing service price, and obtain the optimal crowdsourcing service price.

[0154] Specifically, in this embodiment, the specific functions of the crowdsourcing service price optimization system based on on-demand travel services can also be referred to the corresponding description in the crowdsourcing service price optimization method based on on-demand travel services, and will not be repeated here.

[0155] Based on the above embodiments, the present invention also provides a smart terminal, the principle block diagram of which can be as follows: Figure 4 As shown. This smart terminal can be used to execute the crowdsourcing service price optimization method based on on-demand travel services provided in the above embodiments, which will not be described in detail here for the sake of brevity. The smart terminal includes: a processor coupled to a memory, the memory for storing computer programs or instructions, and the processor for executing the computer programs or instructions stored in the memory, so that the methods in the above method embodiments are executed.

[0156] The present invention also provides a computer-readable storage medium having stored thereon computer instructions for implementing the methods in the above-described method embodiments.

[0157] For example, when the computer program is executed by a computer, it enables the computer to implement the methods described in the above method embodiments.

[0158] This application also provides a computer program product containing instructions that, when executed by a computer, cause the computer to implement the methods described in the above method embodiments.

[0159] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0160] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0161] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0162] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0163] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0164] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for optimizing the price of crowdsourced services based on on-demand transportation services, characterized in that, Includes the following steps: Obtain the initial crowdsourcing service price; Based on the costs and profits of on-demand transportation service providers, a crowdsourced delivery model is constructed. Based on the costs of parcel delivery service providers, construct a multi-warehouse pickup and delivery model; Based on the initial crowdsourcing service price, the crowdsourcing delivery model and the multi-warehouse pickup and delivery model are iteratively solved and optimized using a preset algorithm to obtain the optimal crowdsourcing service price.

2. The method for optimizing the price of crowdsourcing services based on on-demand travel services according to claim 1, characterized in that, The crowdsourced delivery model, based on the costs and profits of on-demand transportation service providers, includes: The total cost of the on-demand mobility service provider is determined based on at least one of the following: pre-acquired fixed operating costs, transportation costs, penalty costs for not meeting passenger travel needs, and passenger compensation costs. The total profit of the on-demand transportation service provider is determined based on the pre-obtained profits from delivering packages and transporting passengers. Based on the total cost and the total profit, a crowdsourced delivery model is constructed.

3. The method for optimizing the price of crowdsourcing services based on on-demand travel services according to claim 1, characterized in that, The aforementioned multi-warehouse pickup and delivery model, based on the costs of parcel delivery service providers, includes: A multi-warehouse pickup and delivery model is constructed based on at least one of the pre-obtained fixed operating costs, transportation costs, and crowdsourced shipping costs of the package delivery service provider.

4. The method for optimizing the price of crowdsourcing services based on on-demand travel services according to claim 1, characterized in that, The step of iteratively solving and optimizing the crowdsourcing delivery model and the multi-warehouse pickup and delivery model based on the initial crowdsourcing service price to obtain the optimal crowdsourcing service price includes: Based on the initial crowdsourcing service price, the crowdsourcing delivery model is solved using a preset first fine-grained tabu search algorithm to calculate the parcel service demand under the initial price; Based on the parcel service demand under the initial price, the multi-warehouse pickup and delivery model is solved using a preset second fine-grained tabu search algorithm to calculate the crowdsourcing service price under the initial parcel delivery demand; The crowdsourcing service price under the initial package delivery demand is updated using a preset genetic algorithm to obtain an optimized crowdsourcing service price; Repeat the steps of calculating the demand for package services and the price of crowdsourcing services, and updating the price of crowdsourcing services, until the preset iteration termination condition is met, and the optimal price of crowdsourcing services is obtained.

5. The method for optimizing the price of crowdsourcing services based on on-demand travel services according to claim 4, characterized in that, Before obtaining the optimal crowdsourcing service price, the following is also included: Obtain passenger travel needs; The initial crowdsourcing service price and the passenger's travel demand are used as inputs to the crowdsourcing delivery model, which outputs the optimal service demand and optimal service route for the on-demand travel service provider.

6. The method for optimizing the price of crowdsourcing services based on on-demand travel services according to claim 5, characterized in that, If the optimal service demand cannot meet the total parcel delivery demand, after outputting the optimal service demand and optimal service route of the on-demand transportation service provider, the following steps are also included: Based on the optimal service demand and the total parcel delivery demand, the remaining parcel delivery demand is determined; Based on the remaining parcel delivery demand, the optimal self-operated service strategy of the parcel delivery service provider is determined.

7. The method for optimizing the price of crowdsourcing services based on on-demand travel services according to any one of claims 4-6, characterized in that, The solution process for the first and second fine-grained tabu search algorithms includes: Obtain passenger travel demand and total parcel delivery demand; The service priority order of the passenger travel demand and the total parcel delivery demand is calculated to obtain the service priority order combination; Based on all combinations of the service sequence, the delivery profit of the package delivery service provider is calculated using the first fine-grained tabu search algorithm, and the delivery cost of the package delivery service provider is calculated using the second fine-grained tabu search algorithm. The crowdsourcing service price is determined based on the delivery profit, the delivery cost, and a preset fine-grained threshold.

8. A crowdsourcing service price optimization system based on on-demand travel services, characterized in that, The system includes: The information acquisition module is used to obtain the initial crowdsourcing service price; The crowdsourced delivery model building module is used to build crowdsourced delivery models based on the costs and profits of on-demand transportation service providers; A multi-warehouse pickup and delivery model building module is used to build multi-warehouse pickup and delivery models based on the costs of parcel delivery service providers; The crowdsourcing service price decision module is used to iteratively solve and optimize the crowdsourcing delivery model and the multi-warehouse pickup and delivery model based on the initial crowdsourcing service price, and obtain the optimal crowdsourcing service price.

9. A smart terminal, characterized in that, include: Memory, used to store executable instructions; A processor for invoking and running the executable instructions in the memory to perform the steps of the crowdsourcing service price optimization method based on any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program instructions that, when executed by a processor, implement the crowdsourcing service price optimization method based on on-demand travel services as described in any one of claims 1-7.

Images