A cooperative model incentive method and system based on Stackelberg game

Abstract

The present invention relates to the field of crowd intelligence perception technology, and in particular to a cooperative mode incentive method and system based on Stackelberg game. The method includes obtaining a task set and user historical data of a crowdsourcing task; calculating the quality score of the crowdsourcing task according to the urgency and energy consumption of the crowdsourcing task; dividing the crowdsourcing task based on the quality score of the crowdsourcing task, and generating cooperative tasks based on the task division; evaluating the ability value of users according to the user historical data; and allocating tasks based on the ability value evaluation results and the generated cooperative tasks, wherein the Stackelberg game is used to obtain the Nash equilibrium solution in the task completion process to obtain the optimal allocation strategy. By introducing the concepts of ability value, shelving level and quality score, the difficult tasks that have existed in the system for a long time are combined, and users are asked to cooperate to complete them, so that the reception completion rate of the tasks is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of group intelligence perception technology, and in particular to a cooperation mode incentive method and system based on Stackelberg game. Background Art

[0002] Mobile crowd sensing is a perception data collection mode based on the participation of large-scale mobile device users. It integrates the perception capabilities of users through sensors of mobile devices such as smartphones and smart watches to achieve extensive, real-time and efficient collection of environmental information. With the popularization of mobile devices and the continuous advancement of technology, mobile crowd sensing has shown great potential in various fields, such as urban transportation, environmental monitoring, emergency response, etc. In the field of collaborative mobile crowd sensing research, a key issue is how to solve the problem of low task coverage. How to motivate users to accept and complete cold tasks in the system is a driving factor for improving the overall task completion rate.

[0003] In the crowd intelligence perception system, there are two roles: request service providers and mobile crowdsourcing users. Request service providers recruit users by publishing tasks to obtain perception data and give users certain incentives; crowdsourcing users receive and complete perception tasks and upload perception data to obtain incentive rewards. However, in reality, when users perform tasks, there are often risks of resource consumption and privacy leakage. Sometimes, the user's resource consumption is not proportional to the reward incentive. There are cases where users consume too many resources to complete a task with very low rewards, resulting in some tasks not being accepted by users for a long time.

[0004] At the same time, there are often some very difficult tasks in the system that require multiple users to complete together, and incentives are allocated according to certain rules. In addition, the incentives given to users by the system are not just rewards. Some incentives include points, experience points, vouchers, etc. The mixed incentive payment method can better stimulate the enthusiasm of users and complete more tasks. Therefore, the incentive allocation of tasks becomes a dimension for users to consider whether to complete the task, and it also becomes an important indicator that needs to be considered in the design of incentive methods, systems and devices. Therefore, the design of incentive methods, systems and devices faces the following challenges:

[0005] (1) The task acceptance rate of some crowd-sensing systems is currently low, and the requesting service providers cannot collect high-quality and complete data in a timely manner, and there is a lack of a method to motivate users to complete tasks. (2) For the calculation of the benefits of both cooperating users, there is a lack of an efficient incentive allocation method to solve the problem under the premise of maximizing user benefits. (3) How to classify the tasks and users in the system and then carry out the next task allocation makes the design of incentive methods, systems and devices more challenging. Therefore, how to motivate users to complete the tasks in the system faster and more comprehensively while ensuring the ideal benefits of users has become a problem that needs to be solved urgently in existing technologies. Summary of the invention

[0006] In order to solve the above-mentioned problems, the present invention provides a cooperative mode incentive method and system based on Stackelberg game. The cold tasks in the system are combined through an improved chaotic particle swarm algorithm (ICPSO) to generate cooperative tasks, which are assigned to users to complete cooperatively, and the weight distribution of rewards and capability values ​​is solved through Stackelberg game to form a Nash equilibrium.

[0007] In a first aspect, the present invention provides a cooperative model incentive method based on Stackelberg game, which adopts the following technical solution:

[0008] A cooperative model incentive method based on Stackelberg game, comprising:

[0009] Obtain the task set and user historical data of crowdsourcing tasks;

[0010] Calculate the quality score of crowdsourcing tasks according to the urgency and energy consumption of crowdsourcing tasks;

[0011] Divide crowdsourcing tasks based on their quality scores, and generate cooperative tasks based on the task division;

[0012] Evaluate the user's capability value based on the user's historical data;

[0013] Task allocation is performed based on the capability evaluation results and the generated cooperative tasks. The Stackelberg game is used to obtain the Nash equilibrium solution in the task completion process and obtain the optimal allocation strategy.

[0014] Furthermore, the quality score of the crowdsourcing task is calculated according to the urgency and energy consumption of the crowdsourcing task, including introducing a sigmoid function to represent the quality score of the task according to the initial urgency and energy consumption of the task, mapping it to the (0, 1) interval, and expressing it in percentage. The calculation formula is:

[0015] = ,

[0016] in Indicates the urgency of the task. represents the energy consumption required for the task. and Represent the weight factors of task urgency and energy consumption respectively, and + =1.

[0017] Furthermore, the crowdsourcing tasks are divided based on the quality scores of the crowdsourcing tasks, including a quality score threshold, and the tasks in the system are divided into high-quality tasks and low-quality tasks, wherein the initial rewards and initial capability values ​​are respectively assigned to the high-quality tasks and the low-quality tasks, and the calculation formulas of the final rewards M and the final capability value V are:

[0018] =

[0019] =

[0020] in, and Represents the basic and basic ability values ​​of the task respectively. and Represents the reward factor and ability value factor of the task, and Expressed as a random reward factor and a random ability value factor.

[0021] Furthermore, the generation of cooperative tasks based on task division includes using an improved chaotic particle swarm optimization algorithm to select high-quality tasks and low-quality tasks to combine and generate cooperative tasks, wherein a holding level L is introduced to determine whether to select tasks for combination, and then the judgment factor choose is updated, and at the same time, a chaotic map is introduced to initialize the position and speed of particles, increase randomness, and expand the diversity of the population, and finally an ICPSO algorithm is applied to generate a task combination plan, calculate the degree of integration of the objective function, and find the global optimal solution.

[0022] Furthermore, the application of ICPSO algorithm to generate task combination scheme, calculate the objective function combination degree, and find the global optimal solution includes solving the objective function combination degree based on the task shelving level gap, the task quality gap, and the similarity of the ratio of the two task rewards and capability values, which is expressed as:

[0023] = + + .

[0024] Furthermore, the user's ability value is evaluated based on the user's historical data, including evaluating the user's ability value based on the number of tasks the user has completed in the past and the quality of the tasks the user has completed in the past. , the calculation formula is:

[0025]

[0026] in and Represents the total number of tasks and the number of tasks completed historically, and Represents the total score of the task and the total score of the tasks completed in the past. and is the capability weight factor, and + = 1.

[0027] Furthermore, the Stackelberg game is used to obtain the Nash equilibrium solution in the process of task completion to obtain the optimal allocation strategy, including respectively calculating the optimal benefits in the reward incentives and capability value incentives of high-ability users and low-ability users, wherein, in the reward incentive allocation, the optimal reward benefit weights of high-ability users and low-ability users are calculated based on the Nash equilibrium solution, thereby obtaining the optimal reward benefits of high-ability users and low-ability users; in the capability value incentive allocation, the optimal capability value benefit weights of high-ability users and low-ability users are calculated based on the Nash equilibrium solution, thereby obtaining the optimal capability value benefits of high-ability users and low-ability users.

[0028] Second, a cooperative model incentive system based on Stackelberg game, including:

[0029] A data acquisition module is configured to acquire a task set and user historical data of a crowdsourcing task;

[0030] The task scoring module is configured to calculate the quality score of the crowdsourcing task according to the urgency and energy consumption of the crowdsourcing task;

[0031] The cooperation module is configured to divide the crowdsourcing tasks based on the quality scores of the crowdsourcing tasks, and generate cooperation tasks based on the task division;

[0032] The capability evaluation module is configured to evaluate the capability value of the user based on the user's historical data;

[0033] The task allocation module is configured to perform task allocation based on the capability value evaluation results and the generated cooperative tasks, wherein the Stackelberg game is used to obtain the Nash equilibrium solution in the task completion process to obtain the optimal allocation strategy.

[0034] In a third aspect, the present invention provides a computer-readable storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device, for example, a method for motivating a cooperative model based on Stackelberg game.

[0035] In a fourth aspect, the present invention provides a terminal device, including a processor and a computer-readable storage medium, the processor being used to implement various instructions; the computer-readable storage medium being used to store multiple instructions, the instructions being suitable for being loaded by the processor and executing the cooperative mode incentive method based on Stackelberg game.

[0036] In summary, the present invention has the following beneficial technical effects:

[0037] (1) The present invention introduces the concepts of capability value, suspension level and quality score, divides tasks into high-quality tasks and low-quality tasks, divides users into high-ability users and low-ability users, generates cooperative tasks, incentivizes high-ability users based on rewards and incentivizes low-ability users based on capability value, combines difficult tasks that have existed in the system for a long time, and allows users to cooperate to complete them, thereby ensuring the completion rate of task acceptance.

[0038] (2) The present invention designs an optimized chaotic particle swarm (ICPSO) algorithm to generate cooperative tasks more effectively. Chaotic mapping is introduced into the algorithm to increase randomness, and the search direction of particles is optimized according to the difference in the task's shelving level, quality gap, incentive similarity and other aspects to avoid particles falling into local optimality. At the same time, task priorities are created according to the shelving levels of high- and low-quality tasks to evaluate the possibility of combining into cooperative tasks. Genetic algorithm mutations are added in the process of updating particle speed and position to improve search accuracy and avoid rapid convergence of solutions, thereby finding the best task combination solution. The results of generating cooperative tasks are optimized and accepted by users as much as possible.

[0039] (3) The present invention applies the theory of Stackelberg game. In the distribution of rewards (ability values), high (low) ability users act as leaders and low (high) ability users act as followers. This is because high-ability users want reward incentives and low-ability users want ability value incentives, but they also want to achieve a state of maximizing benefits in other aspects of incentives. Therefore, the two parties play a game on the distribution weights to form a Nash equilibrium and achieve an incentive distribution that is acceptable to both parties. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a schematic diagram of a cooperative mode incentive method based on Stackelberg game in Example 1 of the present invention;

[0041] Figure 2 It is a schematic diagram of the framework of a cooperative model incentive method based on Stackelberg game in Example 1 of the present invention. DETAILED DESCRIPTION

[0042] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0043] Example 1

[0044] Reference Figure 1 ,The present invention of this embodiment includes two aspects in the first stage of generating cooperative tasks, including a task division stage and a task combination stage;

[0045] Task division stage: The requester uploads the task to the platform system. The system calculates the corresponding quality score based on the initial urgency and energy consumption of the task, and divides it into high-quality tasks and low-quality tasks;

[0046] The specific steps of the task division phase are:

[0047] According to the initial urgency and energy consumption of the task, the sigmoid function is introduced to represent the quality score of the task, mapped to the (0, 1) interval, and expressed as a percentage. The calculation formula is:

[0048] = ,

[0049] in Indicates the urgency of the task. represents the energy consumption required for the task. and Represent the weight factors of task urgency and energy consumption respectively, and + =1.

[0050] Based on quality score threshold , divide the tasks in the system into high-quality tasks and low-quality tasks.

[0051] The initial reward and ability value incentive are given to the task. The higher the difficulty of completing a high-quality task, the more reward and ability value incentive are given. Low-quality tasks are relatively easy to complete, and the reward and ability value incentive are less. The calculation formula is:

[0052] =

[0053] =

[0054] in and They represent the basic incentive reward and basic incentive ability value of the task respectively. and Represents the reward factor and ability value factor of the task.

[0055] and It is expressed as a random reward factor and a random ability value factor, and its value is:

[0056] =

[0057] =

[0058] The system divides the task set into = { } is divided into high-quality tasks and low-quality tasks. We denote the high-quality task set as = { }, the low-quality task set is represented as = { }.

[0059] in Represents the number of tasks in the system, and denote the number of high-quality tasks and low-quality tasks respectively, and .

[0060] Each task is represented as ={ , , , , , }, the attributes are The urgency, energy consumption, shelving level, incentive reward, incentive capacity value and quality score of each task.

[0061] Task combination stage: In order to recruit more suitable users, we apply the improved chaotic particle swarm optimization (ICPSO) algorithm to select high-quality tasks and combine them with low-quality tasks to generate cooperative tasks.

[0062] The specific steps of the task integration phase are:

[0063] Introducing shelving levels , determine whether to select this task for combination, and then update the judgment factor , expressed as:

[0064] =

[0065] in Indicates the hold threshold for selecting tasks for combining.

[0066] Chaotic mapping is introduced to initialize the position and velocity of particles, increase randomness, and expand the diversity of the population.

[0067] Apply the ICPSO algorithm to generate a task combination plan, calculate the degree of integration of the objective function, and find the global optimal solution. Representation:

[0068]

[0069] ,

[0070] in Indicates the shelving level of high-quality tasks, Indicates the shelving level of low-quality tasks.

[0071] Among them, the first factor affecting the task integration is the task shelving level difference. Its function is expressed as:

[0072] =

[0073] in is the regulating factor of the function.

[0074] The second influencing factor is the quality gap of the task. The quality gap function is expressed as:

[0075] =

[0076] in is the quality score of the low-quality task, is the quality score of high-quality tasks.

[0077] The third factor affecting the evaluation of task integration is the similarity of the ratio of the two task rewards and the ability value incentives, and its function formula is:

[0078] =

[0079] in =( ), =( ), and Represent the incentive reward and capability value of low-quality tasks respectively, and They represent the incentive reward and ability value of high-quality tasks respectively.

[0080] Based on the above discussion, The formula is expressed as:

[0081] = + +

[0082] The algorithm iterates and generates a set of cooperative tasks = .

[0083] The attributes of the cooperative task only include rewards and ability value incentives. The rewards and ability value incentives are respectively and To express.

[0084] The second stage of the Stackelberg game between users of the present invention includes two aspects, including a user division stage and a Stackelberg solving stage;

[0085] The steps in the user segmentation phase are:

[0086] Evaluate the user's ability value based on the number of tasks they have completed and the quality of their completed tasks , the calculation formula is:

[0087]

[0088] in and Represents the total number of tasks and the number of tasks completed historically, and Represents the total score of the task and the total score of the tasks completed in the past. and is the capability weight factor, and + = 1.

[0089] Set a capability threshold , divide the users in the system into high-ability users and low-ability users.

[0090] The ability value of high-ability users is To represent, the ability value of low-ability users is expressed by To express.

[0091] Initial set of users The user set is divided into a high-capability user set and a low-capability user set. The high-capability user set is represented as , the low-ability user set is represented as .

[0092] in represents the number of initial users, and are respectively represented by the number of high-capability users and the number of low-capability users divided, and .

[0093] Solving the Stackelberg stage: After generating cooperative tasks, recruit suitable high-ability users and low-ability users to perform tasks. High-ability users rely on their own ability advantages to obtain more reward incentives and are motivated to participate in cooperative tasks. Low-ability users want to obtain more ability points to improve their abilities, so that they can participate in higher-quality tasks, so they are motivated to participate in cooperative tasks. However, high-ability users do not only consider the reward factor, but also want to obtain higher ability value incentives. At the same time, low-ability users may only want ability points and also want to obtain more reward incentives.

[0094] Therefore, the Stackelberg game is introduced. In the reward incentive game, high-ability users act as leaders and propose their own distribution weights, while low-ability users act as followers and refer to the strategies of high-ability users to consider their own optimal benefit weights. Finally, the Nash equilibrium state is reached, and the distribution weight strategy that neither party will change is obtained.

[0095] The specific solution steps are:

[0096] In order to motivate users to complete cooperative tasks more efficiently, additional rewards and ability value incentives that decrease over time are introduced. The shorter the time it takes for users to complete tasks, the more additional incentives they will receive, and vice versa.

[0097] The reward incentive for cooperative tasks is as follows:

[0098] = +

[0099] in The basic reward for cooperative tasks. As additional incentive compensation.

[0100] The basic reward is the sum of the reward incentives for high-quality tasks and low-quality tasks, expressed as:

[0101] = +

[0102] The additional incentive compensation will decrease as the time it takes for high-ability users and low-ability users to complete the task increases, expressed as:

[0103] = *

[0104] in and Respectively represent the time that high-ability users and low-ability users spend completing their own parts of the collaborative task, It is an additional incentive regulation factor.

[0105] The reward allocation weight strategy function for high-capability users is expressed as , the reward allocation weight strategy function for low-ability users is expressed as , the allocation strategy function expressions of both parties are as follows:

[0106] = + -

[0107] = + -

[0108] in is the base weight assigned, and is the regulating factor of the function.

[0109] Allocation weight strategy function for low-ability users In Taking partial derivative, we get :

[0110] = -

[0111] Let the partial derivative equal to zero and solve the following solution for low-ability users :

[0112] =

[0113] Will be obtained Allocation strategy for bringing in high-capability users In, we get:

[0114] = + -

[0115] For Taking partial derivatives, we get :

[0116] = + -

[0117] Set it equal to zero to obtain the optimal completion time of the optimal allocation weight for high-capability users :

[0118] =

[0119] Will Bring in the follow-up solution Find the optimal completion time for the optimal allocation weights of low-ability users :

[0120] =

[0121] The optimal time Nash equilibrium solution of Stackelberg game , the optimal benefit weights of high-capacity users and low-capacity users and It can be expressed as the following formula:

[0122] =

[0123] =

[0124] Get the best compensation benefits for high-capacity users and low-capacity users and :

[0125] = *

[0126] = *

[0127] In the Stackelberg game in terms of capability value, low-ability users are leaders and high-ability users are followers. Low-ability users first set their own allocation weight strategy, and high-ability users set their own optimal response strategy based on known information.

[0128] The definition of the ability value incentive of the cooperative task is similar to the reward incentive of the above cooperative task, which is expressed as:

[0129] = +

[0130] in To encourage cooperative task ability value, It is expressed as the basic ability value incentive for cooperative tasks. Provides additional ability value incentives.

[0131] = + , = *

[0132] in Expressed as the ability value incentive for high-quality tasks, Represents the ability value incentive for low-quality tasks.

[0133] Similar to the reward Stackelberg game, the ability value allocation weight strategy for low-ability users is , the ability value allocation weight strategy for high-ability users is , the expression is as follows:

[0134] = + -

[0135] = + -

[0136] After the same Stackelberg game process as above, the same Nash equilibrium solution will be obtained in the end .

[0137] The optimal ability value benefit weights of high-ability users and low-ability users and It is expressed as:

[0138] =

[0139] =

[0140] The optimal capability value benefits of high-capability users and low-capability users are expressed as and ,get:

[0141] = *

[0142] = *

[0143] In order to verify the performance of the ICPSO algorithm, the ICPSO algorithm was experimentally compared with the method without adding cooperation mode, the task bundling method based on genetic algorithm (GA), the task combination method based on simulated annealing (SA), and the optimization method based on Harris Eagle algorithm on MovieLens and Book-Crossing datasets. The experimental results show that the performance of the ICPSO algorithm is better than other comparison methods in terms of task completion rate, average user reward income, average user ability value income, and average user energy consumption, thus verifying the superiority and effectiveness of the ICPSO algorithm.

[0144] As a further embodiment,

[0145] This embodiment provides a cooperative mode incentive method based on Stackelberg game, such as Figure 1 As shown, the specific steps are as follows:

[0146] Step 1: Task division. The service requester will upload several tasks to the MCS system, and the system will divide the tasks into high-quality tasks and low-quality tasks based on the quality scores of the tasks, so as to recruit suitable users for assignment.

[0147] Step 2: Generate cooperative tasks. In order to better solve the problem of task shelving, the system will apply the improved chaotic particle swarm optimization (ICPSO) algorithm to combine tasks that have not been accepted by users for a long time according to certain standards, generate cooperative tasks, and motivate users in the system to cooperate to complete them.

[0148] Step 3: User segmentation. The users in the system will be divided into high-ability users and low-ability users according to their ability values, and the cooperative tasks will be matched to the appropriate combination of high-ability users and low-ability users, so as to improve the task completion rate in the system through user cooperation.

[0149] Step 4: Game-based allocation: When high-ability users and low-ability users collaborate to complete a task, both parties will play a Stackelberg game to obtain rewards and ability value allocation based on their own ability values ​​and completion time.

[0150] Figure 2 This is the method flow chart of this method. The specific process is as follows:

[0151] In the first stage of the method, the service requester uploads the task to the system. The system evaluates the difficulty of the task based on the initial urgency and energy consumption of the task and introduces a quality score to represent it.

[0152] The system divides tasks into high-quality tasks and low-quality tasks based on the differences in task quality scores, and selects tasks that are greater than the shelving threshold, that is, a combination of high-quality tasks and low-quality tasks that remain in the system for a long time.

[0153] The improved chaotic particle swarm optimization (ICPSO) algorithm is used to generate cooperative tasks. Chaotic mapping is introduced to evaluate whether tasks should be combined based on the difference in the shelving levels, quality score differences and incentive similarity between low-quality tasks and high-quality tasks to generate cooperative tasks.

[0154] In the second stage, the system evaluates the user's ability value based on the number of tasks completed in the past and the quality score of the tasks completed in the past, and divides the user into high-ability users and low-ability users according to the user's ability value.

[0155] The system selects appropriate high-ability users and low-ability users to complete the cooperative task. During the cooperation process, both parties will maximize their own benefits and play the Stackelberg game.

[0156] In the reward-benefit game, high-ability users act as leaders and low-ability users act as followers; in the ability-value-benefit game, low-ability users act as leaders and high-ability users act as followers.

[0157] Under the state of Nash equilibrium of the game, the optimal task completion time for high-ability users and low-ability users is finally obtained, thereby obtaining the optimal profit ratio for both parties.

[0158] In order to verify the effectiveness of the method in this embodiment, the ICPSO algorithm was experimented with real data sets. The data sets are movie rating data sets (MovieLens) and book rating data sets (Book-Crossing). The above real data sets are derived from existing public data and have obtained the consent of the information owner.

[0159] Example 2

[0160] This embodiment provides a cooperation mode incentive system based on Stackelberg game.

[0161] A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device, for a cooperative mode incentive method based on Stackelberg game.

[0162] A terminal device includes a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, wherein the instructions are suitable for being loaded and executed by the processor to implement a cooperative mode incentive method based on Stackelberg game.

[0163] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A cooperative mode incentive method based on Stackelberg game, applied to the crowd intelligence perception system, characterized in that: include: Obtaining the task set and user historical data of the crowdsourcing task; the requester uploads the crowdsourcing task to the platform system; Calculate the quality score of crowdsourcing tasks according to the urgency and energy consumption of crowdsourcing tasks; Divide crowdsourcing tasks based on their quality scores, and generate cooperative tasks based on the task division; Evaluate the user's capability value based on the user's historical data; Task allocation is performed based on the ability value evaluation results and the generated cooperative tasks. The Stackelberg game is used to find the Nash equilibrium solution in the task completion process and obtain the optimal allocation strategy. This includes calculating the optimal benefits of reward incentives and ability value incentives for high-ability users and low-ability users respectively. The crowdsourcing tasks are divided based on the quality scores of the crowdsourcing tasks, including the quality score threshold, and the tasks in the system are divided into high-quality tasks and low-quality tasks, wherein the initial rewards and initial capability values ​​are respectively given to the high-quality tasks and the low-quality tasks, and the calculation formulas of the final rewards M and the final capability value V are: in, and Represents the basic reward and basic ability value of the task respectively. and Represents the reward factor and ability value factor of the task, and Expressed as a random reward factor and a random ability value factor; The generation of cooperative tasks based on task division includes using an improved chaotic particle swarm optimization algorithm to select high-quality tasks and low-quality tasks to generate cooperative tasks, wherein a shelving level L is introduced to determine whether to select tasks for combination, and then the judgment factor choose is updated, and at the same time, a chaotic map is introduced to initialize the position and speed of particles, increase randomness, and expand the diversity of the population, and finally an ICPSO algorithm is applied to generate a task combination plan, calculate the degree of combination of the objective function, and find the global optimal solution; The ICPSO algorithm is applied to generate a task combination scheme, calculate the objective function binding degree, and find the global optimal solution, including solving the objective function binding degree based on the task shelving level gap, the task quality gap, and the similarity of the ratio of the two task rewards and capability values, which is expressed as: ； in, Represents the task shelving level gap, represents the quality gap of the task, Represents the similarity of the ratio of the rewards and ability values ​​of two tasks; specifically, and They are respectively represented as the shelving levels of low-quality tasks and high-quality tasks, and are respectively the quality scores of low-quality tasks and high-quality tasks, and , and Represent the incentive reward and capability value of low-quality tasks respectively, and They represent the incentive reward and capability value of high-quality tasks respectively; The user's ability value is evaluated based on the user's historical data, including evaluating the user's ability value based on the number of tasks the user has completed in the past and the quality of the tasks the user has completed in the past. , the calculation formula is: in and Represents the total number of tasks and the number of tasks completed historically, and Represents the total score of the task and the total score of the tasks completed in the past. and is the capability weight factor, and .

2. A method for motivating a cooperative model based on Stackelberg game according to claim 1, characterized in that: The quality score of the crowdsourcing task is calculated according to the urgency and energy consumption of the crowdsourcing task, including introducing a sigmoid function to represent the quality score of the task according to the initial urgency and energy consumption of the task, mapping it to the (0, 1) interval, and expressing it in percentage. The calculation formula is: in Indicates the urgency of the task. represents the energy consumption required for the task. and Represent the weight factors of task urgency and energy consumption respectively, and .

3. A method for motivating a cooperative model based on Stackelberg game according to claim 2, characterized in that: The method of respectively calculating the optimal benefits in the reward incentive and capability value incentive of high-ability users and low-ability users includes, in the reward incentive allocation, respectively calculating the optimal reward benefit weights of high-ability users and low-ability users based on the Nash equilibrium solution, thereby obtaining the optimal reward benefits of high-ability users and low-ability users; in the capability value incentive allocation, respectively calculating the optimal capability value benefit weights of high-ability users and low-ability users based on the Nash equilibrium solution, thereby obtaining the optimal capability value benefits of high-ability users and low-ability users.

4. A cooperative mode incentive system based on Stackelberg game, executing the method as claimed in claim 1, characterized in that: include: A data acquisition module is configured to acquire a task set and user historical data of a crowdsourcing task; The task scoring module is configured to calculate the quality score of the crowdsourcing task according to the urgency and energy consumption of the crowdsourcing task; The cooperation module is configured to divide the crowdsourcing tasks based on the quality scores of the crowdsourcing tasks, and generate cooperation tasks based on the task division; The capability evaluation module is configured to evaluate the capability value of the user based on the user's historical data; The task allocation module is configured to perform task allocation based on the capability value evaluation results and the generated cooperative tasks, wherein the Stackelberg game is used to obtain the Nash equilibrium solution in the task completion process to obtain the optimal allocation strategy.

5. A computer-readable storage medium storing a plurality of instructions, characterized in that: The instructions are suitable for being loaded by a processor of a terminal device and executing the method according to claim 1 .

6. A terminal device, comprising a processor and a computer-readable storage medium, wherein the processor is used to implement each instruction; and the computer-readable storage medium is used to store multiple instructions, characterized in that: The instructions are suitable for being loaded by a processor and executing the method as claimed in claim 1 .

Images