A high-dimensional optimization task allocation method and system considering three-party multi-objectives

Abstract

The present invention relates to the intersection of crowd intelligence perception and artificial intelligence technology, and in particular to a high-dimensional optimization task allocation method and system that considers three-party multi-objectives. The method includes obtaining a set of attributes corresponding to crowdsourcing workers and crowdsourcing tasks; constructing an optimization problem for a multi-party scenario based on the obtained attribute set, including constructing an optimization problem for platform profit; establishing an optimization problem for maximizing worker benefits; establishing an optimization problem for maximizing task quality robustness; establishing an optimization problem for maximizing requester benefits; and optimizing the optimization problem for the multi-party scenario using an improved RVEA optimization algorithm to obtain an optimization result. The task allocation method proposed in the present invention achieves better allocation results in terms of allocation efficiency and allocation utility; its allocation results are superior to those obtained by other single-objective optimization algorithms in terms of service quality and allocation efficiency.

Description

Technical Field

[0001] The present invention relates to the intersection of crowd intelligence perception and artificial intelligence technology, and in particular to a high-dimensional optimization task allocation method and system that considers three-party multi-objectives. Background Art

[0002] With the advancement of mobile devices such as smartphones, mobile crowdsensing has become a new cost-effective perception paradigm. Mobile users can use their devices to upload perception information such as air quality, traffic conditions, and environmental images. With the growth of data and the increase in dimensionality, high-dimensional and large-scale tasks are the key issues that need to be studied in task allocation. In the field of high-dimensional optimization task allocation, considering three-party multi-objectives is a complex and challenging task. Such problems usually involve multiple decision variables and multiple decision objectives, each of which may have contradictory situations.

[0003] High-dimensional optimization task allocation often involves multiple decision-making objectives, which may involve different performance indicators or concerns. Traditional single-objective optimization problems cannot meet the needs of multi-objective task allocation, so multi-objective optimization techniques are needed. For example, multi-objective genetic algorithms, multi-objective particle swarm algorithms, etc. are used to deal with the trade-offs between different objectives.

[0004] In task allocation, conflicts and competitions among three parties are usually involved. This may include competition for resources, task priorities, and trade-offs between different goals. To solve this situation, we need to consider game theory, collaborative optimization, and other methods to ensure that the cooperation and competition among the three parties can be balanced. Therefore, on this basis, in order to effectively improve the efficiency and effectiveness of allocation, we still face the following difficulties: (1) When it comes to cooperation and competition among multiple parties, the system often becomes very complex. It is a challenge to correctly model the system and design appropriate game theory or collaborative optimization algorithms to deal with this complexity. (2) In practical applications, efficient computing and algorithms are required to handle large-scale game and collaborative optimization problems. Considering computational complexity and algorithm efficiency are key factors in the solution. (3) Task allocation solutions that only consider a single goal will result in a narrow scope of solution and are prone to falling into local optimal solutions. The goal of maximizing long-term social welfare cannot be achieved. (4) The dynamic changes and uncertainties of the environment may affect game and collaborative optimization. Effective solutions need to take these changes and uncertainties into account and be adaptable. (5) The task allocation scheme that considers multiple objectives serially cannot find the optimal equilibrium solution, but instead obtains different optimal equilibrium solutions. Therefore, it is necessary to consider the task allocation scheme of multiple objectives in parallel. Summary of the invention

[0005] In order to solve the above-mentioned problems, the present invention provides a high-dimensional optimization task allocation method and system that considers three-party multi-objectives. The method is based on a high-dimensional multi-objective optimization method that balances the needs of the requester, the platform, and the workers to balance the cost of the workers, the profit of the platform, and the robustness and service quality of the task. A collaborative optimization algorithm is designed so that the three parties can work together to jointly optimize the goals of the task. In order to maximize the overall utility, a dynamic adaptability mechanism is introduced so that the system can dynamically adjust the target weights and task allocation strategies to adapt to the dynamic changes in the task environment.

[0006] In the first aspect, the present invention provides a high-dimensional optimization task allocation method considering three-party multi-objectives, which adopts the following technical solution:

[0007] A high-dimensional optimization task allocation method considering tripartite multi-objectives, comprising:

[0008] Get the attribute set corresponding to crowdsourcing workers and crowdsourcing tasks;

[0009] Based on the acquired attribute set, we construct optimization problems for multi-party scenarios, including optimization problems for platform profits, optimization problems for maximizing worker benefits, optimization problems for maximizing task quality robustness, and optimization problems for maximizing requester benefits.

[0010] The improved RVEA optimization algorithm is used to optimize the optimization problem of multi-party scenarios and obtain the optimization results.

[0011] Furthermore, the method of obtaining the attribute sets corresponding to crowdsourcing workers and crowdsourcing tasks includes collecting the attribute sets corresponding to crowdsourcing workers and crowdsourcing tasks, and extracting the social attributes of the workers to obtain the social attributes of the crowdsourcing workers in each spatiotemporal region; considering the location and equipment requirement attributes of the tasks at the same time; saving the real-time attributes of the workers and tasks, and generalizing the privacy location attributes of the workers.

[0012] Furthermore, the optimization problem of constructing the platform profit includes establishing the optimization problem according to the social attributes of the crowdsourcing workers, the social attributes of the crowdsourcing tasks themselves, and the platform profit; taking the platform profit as the optimization problem of the platform, while considering the social attributes of the workers and the tasks, the objective function of the optimization problem is expressed as:

[0013]

[0014] Among them, the social attribute of worker i is represented by W i The social attribute of task j is represented by T j The platform profit is represented by P, and the task allocation is represented by a binary variable X ij , where X ij= 1 if worker i is assigned to task j, otherwise 0; P ij is the payoff of worker i for performing task j.

[0015] Furthermore, the optimization problem of maximizing worker benefits is established, including considering the benefits of each crowdsourcing worker performing tasks, the degree of adaptability of the social attributes of crowdsourcing workers to crowdsourcing tasks, and the profit of the platform. Suppose there are N workers, M tasks, and the benefit of worker i performing task j is C w , the worker's cost is cost, and the social attribute of worker i is W i , the social attribute of task j is T j , the profit of the platform is P, and the objective function of the optimization problem is expressed as:

[0016] .

[0017] Furthermore, the optimization problem of maximizing the robustness of task quality is established, including considering the quality of each crowdsourcing task, the adaptability of crowdsourcing workers to crowdsourcing tasks, and the evaluation of workers by the quality of tasks. There are N workers, M tasks, and the quality of worker i performing task j is Q ij , the social attribute of worker i is W i , the social attribute of task j is T j ; The goal is to find a task allocation scheme that maximizes the robustness of task quality and makes the quality fluctuation of all tasks small. The objective function of the optimization problem is expressed as:

[0018]

[0019] Among them, Q j is the average quality of task j, and Q is the average quality of all tasks.

[0020] Furthermore, the optimization problem of maximizing the benefit of the requester is established, including considering the quality of each crowdsourcing task, the completion time of the task, the evaluation of the crowdsourcing workers, and the importance of the requester to the task. The objective function of the optimization problem is expressed as:

[0021]

[0022] Among them, U ij is the benefit obtained by the requester from the completion of task j.

[0023] Furthermore, the improved RVEA optimization algorithm is used to optimize the optimization problem of the multi-party scenario, including calculating the Pareto optimal solution for the objective function of the multi-party scenario based on the RVEA multi-objective optimization algorithm.

[0024] In the second aspect, a high-dimensional optimization task allocation system considering three-party multi-objectives includes:

[0025] The data acquisition module is configured to obtain a set of attributes corresponding to crowdsourcing workers and crowdsourcing tasks;

[0026] The optimization problem building module is configured to build optimization problems of multi-party scenarios based on the acquired attribute set, including building an optimization problem of platform profit; building an optimization problem of maximizing worker benefits; building an optimization problem of maximizing task quality robustness; and building an optimization problem of maximizing requester benefits.

[0027] The optimization module is configured to optimize the optimization problem of the multi-party scenario using the improved RVEA optimization algorithm to obtain the optimization result.

[0028] 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, a high-dimensional optimization task allocation method considering three-party multi-objectives.

[0029] In a fourth aspect, the present invention provides a terminal device comprising 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 method for high-dimensional optimization task allocation that considers three-party multi-objectives.

[0030] In summary, the present invention has the following beneficial technical effects:

[0031] (1) The task allocation method proposed in the present invention achieves better allocation results in terms of allocation efficiency and allocation utility; its allocation results are superior to those obtained by other single-objective optimization algorithms in terms of service quality and allocation efficiency.

[0032] (2) The task allocation method of the present invention maintains the robustness of the algorithm while guiding the search space through RVEA, which means that it can effectively find high-quality solutions in the face of the diversity and complexity of problems.

[0033] (3) The task allocation method of the present invention can better balance the interests and goals of different parties by comprehensively considering the goals of multiple parties. By considering more participants, the needs of all parties can be more fully understood and met, thereby improving overall benefits. The three-party task allocation can better avoid or mitigate possible conflicts of interest. By comprehensively considering the goals of different parties, a more consensus-based task allocation plan can be formulated to reduce potential friction and conflict. By coordinating and optimizing the three-party task allocation, the entire system can be optimized and the overall efficiency can be improved. This is particularly important for large-scale, complex projects or systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a description diagram of task allocation in the opportunistic MCS scenario of the present invention.

[0035] Figure 2 It is a schematic diagram of the process of the present invention.

[0036] Figure 3 This is an illustration of the multi-objective optimization algorithm used in the present invention.

[0037] Figure 4 This is a comparison chart of the effects of the comprehensive evaluation index considered by the present invention and other methods.

[0038] Figure 5 This is a service quality comparison diagram of the task allocation achieved by the present invention and other optimization algorithms.

[0039] Figure 6 This is a comparison diagram of the social welfare values ​​of task allocation achieved by the present invention and other optimization algorithms.

[0040] Figure 7 This is a comparison diagram of worker return values ​​for task allocation achieved by the present invention and other optimization algorithms.

[0041] Figure 8 This is a comparison diagram of worker satisfaction between the present invention and other optimization algorithms for task allocation. DETAILED DESCRIPTION

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

[0043] Example 1

[0044] refer to Figure 1 , Figure 1 It is a task allocation scenario diagram in a real MCS scenario. In the mobile crowdsourcing system, there are three main roles: crowdsourcing workers, crowdsourcing tasks, and crowdsourcing platforms; task publishers publish tasks to crowdsourcing platforms or specific mobile applications. When publishing tasks, task publishers usually need to set the reward and completion deadline of the task, as well as the skills and conditions required for the task. The mobile crowdsourcing platform will match the task with suitable mobile device users. This matching process may involve factors such as the user's geographical location, skills, and historical performance. Task allocation algorithms can be used to determine the best task distribution strategy. Once mobile device users accept the task, they start to perform the task. Crowdsourcing platforms usually monitor the execution progress of the task in real time to ensure that the task is completed on time and meets the requirements. In order to ensure the quality of the task results, crowdsourcing platforms may take some quality control measures, such as review, voting mechanism, or algorithm-based quality assessment. Mobile crowdsourcing platforms usually provide evaluation and feedback mechanisms between users to record users' performance and reputation. This can help improve the efficiency and quality of task distribution.

[0045] Please refer to Figure 2 , Figure 2 : is a flow chart of the method of the present invention; the present invention provides a multi-objective task allocation method based on spatiotemporal crowdsourcing, which specifically includes the following steps:

[0046] S1. Collect the attribute sets corresponding to crowdsourcing workers and crowdsourcing tasks, and extract the social attributes of the workers to obtain the social attributes of crowdsourcing workers in each spatiotemporal region;

[0047] In step S1, the social attributes include: the traffic conditions of the roads where the crowdsourcing workers perform their work and the route planning conditions of the mutual influence between multiple pedestrians. At the same time, the equipment attributes and capability attributes of the workers are also obtained to ensure the quality of the workers' completion of the tasks.

[0048] As an embodiment, specifically, in step S1, crowdsourcing worker W i The attribute set is <l wi ,b wi, r wi, k wi >, where l wi For crowd workers in b wi The position of the crowdsourcing worker when entering the crowdsourcing platform at time l wi Change with the workers' movement; wi is the worker's ability attribute value, indicating the worker's proficiency and experience in performing tasks; r wi It is the equipment attribute of the worker. The equipment attribute of the worker must meet the requirements of different tasks. Specific tasks require specific equipment attributes of the worker.

[0049] In step S1, crowdsourcing task t i The attribute set is , where l ti For crowdsourcing tasks in b ti The starting position of the crowdsourcing task when entering the crowdsourcing platform at that moment and l ti There will be no change; ti The shortest deadline to complete the task, the task must be completed within c ti The task must be completed within the time limit, otherwise it will affect the quality of the task. B is the budget of the task. The requester provides different budgets according to the different needs of his task. It is the equipment attribute of the task, indicating that the equipment requirement attribute of the task must be met by the worker to perform the task. The corresponding attributes need to be matched.

[0050] S2. Construct an optimization problem based on the social attributes of the workers submitted by the workers, the social attributes of the tasks themselves, and the profits of the platform; and treat the platform profits as the optimization problem of the platform.

[0051] S3. Construct an optimization problem to maximize worker benefits; consider the following factors: the benefits of each worker performing the task, the degree of adaptability of the worker's social attributes to the task, and the platform's profit.

[0052] S4. Establish an optimization problem to maximize the robustness of task quality; consider the following factors: the quality of each task, the worker's adaptability to the task, and the evaluation of the worker by the quality of the task. There are N workers and M tasks. The quality of worker i performing task j is Qij, the social attribute of worker i is Wi, and the social attribute of task j is Tj. The goal is to find a task allocation scheme that maximizes the robustness of task quality, that is, to make the quality fluctuation of all tasks small.

[0053]

[0054] Where Qj is the average quality of task j and Q is the average quality of all tasks.

[0055] The optimal allocation solution specifically includes a balanced solution for conflicting indicators of cost and service quality.

[0056] The qualities of service for the described tasks are as follows:

[0057]

[0058] In the above formula, Uw is the worker's willingness to perform the task; the worker's willingness to perform the task is related to the task and the worker's location. The farther the location is, the less willing the worker is to perform the task. Therefore, it is necessary to calculate the distance between the worker and the task location, so it is necessary to predict the worker's movement route when calculating the worker's willingness.

[0059] rek is the worker’s ability to complete tasks measured by the number of tasks performed by the worker and the number of valid tasks; It is a time function to calculate the impact of task completion time on task service quality. ϕ i = min [ 1 − max( 0 ,min [ log v ( Ct ), 1 ] ) ]

[0060] The cost of the platform is as follows:

[0061] COST=∑red j +cos tp

[0062] In the above formula, j is the number of workers performing the task, and the value of j is 1, 2, 3…m; red j =a0+D(l ti ,l wj )*a+c j is the formula for calculating workers' compensation, where a0 is the basic reward, D(l ti ,lwj )*a is the scheduling reward calculated by the task location and worker location, c j Cumulative rewards for completing tasks; cos tp Represents the daily maintenance cost of the platform.

[0063] S5. Establish an optimization problem to maximize the benefit of the requester; consider the following factors: the quality of each task, the completion time of the task, the worker's evaluation, and the importance of the task to the requester.

[0064]

[0065] Among them, U ij is the benefit that the requester gets from the completion of task j. This benefit can be defined based on the task quality, completion time, worker evaluation, and task importance. Give a task allocation and scheduling scheme that maximizes the total benefit of the requester.

[0066] S6. Consider the decision area space of parallel multi-objective optimization of multiple demands, and find the Pareto optimal solution set that balances the multiple demand indicators in the feasible area. The present invention will use the improved RVEA optimization algorithm to optimize the optimization problems of complex multi-party scenarios in reality.

[0067] Resource constraints: Consider the resources involved in task allocation, such as time, money, manpower, etc. Ensure that the task allocation plan is feasible in terms of resources and does not exceed the system's tolerance limits.

[0068] Capability and equipment matching constraints: Ensure that the skills of the executors assigned to the task match the requirements of the task. This can be achieved by defining a skill matching matrix or a skill matching function to ensure that each executor has the ability to complete the assigned task.

[0069] Inter-task constraints: Consider the correlation and dependencies between tasks. Some tasks may need to be started after other tasks are completed, or there may be a certain order between them. These constraints can be represented by a task graph or a task dependency matrix.

[0070] Conflict constraints: If there are conflicts between performers, ensure that the task allocation scheme does not lead to escalation of conflicts. This may require taking into account the relationships, preferences, and possible competition between performers.

[0071] The multi-objective optimization, the present invention adopts RVEA (Reference Vector Guided Evolutionary Algorithm), which is a multi-objective optimization algorithm for solving problems involving multiple objective functions. When there are two or more optimized objective functions, it is called multi-objective optimization. The solution of multi-objective is usually a set of equilibrium solutions. In multi-objective optimization, there are multiple optimization objectives, so there is no longer a single optimal solution. On the contrary, there is a set of solutions, called Pareto optimal solutions or non-dominated solution sets. Pareto optimal solution set is an important concept in multi-objective optimization problems, which refers to a set of solutions that cannot be dominated by other solutions in the objective space. The Pareto optimal solution set shows a series of solutions that achieve the best performance in multiple objectives. These solutions may be relatively good in some objectives and relatively poor in other objectives, forming a balance. In the Pareto optimal solution set, each solution is non-dominated. This means that no other solution can be better than it in all objectives at the same time, or for any other solution, at least one objective is not as good as this solution. The Pareto optimal solution set constitutes the Pareto frontier, that is, a set of non-dominated solutions in the objective space. This frontier represents the feasible and excellent solution space of the problem. The Pareto optimal solution set usually shows diversity because it contains solutions that have advantages in different objectives. This helps to provide decision makers with a variety of choices so that they can adjust according to actual needs. The Pareto optimal solution set shows the trade-offs and balances between multiple objectives. Each solution in the solution set is obtained by weighing different objectives. There is no absolute "best" solution, but a result of a trade-off. The RVEA algorithm includes the following:

[0072] 1. Initialize the population Generate an initial population, usually by randomly generating multiple individuals, each of which represents a potential solution (a matching pair between a worker and a task). 2. Generate reference vectors Create uniformly distributed reference vectors to guide population evolution. Reference vectors are usually distributed in the target space, combining the tripartite benefits of the requester, the platform, and the worker to optimize the objective function. 3. Evaluate fitness Calculate the fitness value of each individual under all objective functions. The fitness value is used to evaluate the quality of individuals. 4. Environmental selection Select individuals with better fitness as part of the next generation population. RVEA uses an angle-based selection mechanism to select individuals based on their angle with the reference vector. 5. Mutation and crossover operations Mutate and crossover operations on the selected individuals to generate new individuals. Mutation and crossover operations increase the diversity of the population and help search for better solutions. 6. Update the population Add the newly generated individuals to the population and re-evaluate the fitness of all individuals. According to the fitness and angle selection mechanism, retain good individuals and eliminate poor individuals. 7. Dynamically adjust the reference vector According to the evolution of the population, dynamically adjust the position of the reference vector so that the reference vector can better guide the population to search the Pareto frontier. 8. Repeat the evolution Repeat steps 4 to 7 until the preset stop condition is reached, such as the number of iterations reaches the upper limit or the population changes tend to be stable. 9. Output the results Output the final population, and the individuals in the population are the approximate Pareto optimal solution set. The individuals in the population are the worker and task matching sequences that meet different optimization goals.

[0073] Finally, please refer to Figure 5 , Figure 5 It is a comparison chart of the effects of the comprehensive evaluation index considered in the present invention and other methods; the present invention experiments on a real data set of multi-party task allocation problems in real mobile crowdsourcing scenarios, and visualizes the solution sets therein.

[0074] Figure 5 It shows that the adaptive high-dimensional optimization algorithm proposed in this paper can effectively improve the quality of service (QoS). Compared with other optimization algorithms, the method proposed in this invention is superior to other comparative algorithms. Figure 6 It shows that the adaptive high-dimensional optimization algorithm proposed in this paper can consider the different benefits of three parties to improve the comprehensive social welfare value. Compared with other optimization algorithms, the method proposed in this invention is superior to other comparative algorithms. Figure 7 It is shown that the worker return value of the method proposed in the present invention is better than that of other comparison algorithms in multiple rounds of experiments, which shows that the method proposed in the present invention can effectively improve the benefits of workers. Figure 8 It is shown that the method proposed in the present invention has better worker satisfaction than other comparison algorithms in multiple rounds of experiments, which shows that the method proposed in the present invention can effectively improve the efficiency of workers.

[0075] Example 2

[0076] This embodiment provides a high-dimensional optimization task allocation system considering three parties and multiple objectives, including:

[0077] The data acquisition module is configured to obtain a set of attributes corresponding to crowdsourcing workers and crowdsourcing tasks;

[0078] The optimization problem building module is configured to build optimization problems of multi-party scenarios based on the acquired attribute set, including building an optimization problem of platform profit; building an optimization problem of maximizing worker benefits; building an optimization problem of maximizing task quality robustness; and building an optimization problem of maximizing requester benefits.

[0079] The optimization module is configured to optimize the optimization problem of the multi-party scenario using the improved RVEA optimization algorithm to obtain the optimization result.

[0080] 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 high-dimensional optimization task allocation method considering tripartite multi-objectives.

[0081] 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 by the processor and executing the high-dimensional optimization task allocation method considering three-party multi-objectives.

[0082] 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 high-dimensional optimization task allocation method considering three-party multi-objectives, characterized in that: include: Get the attribute set corresponding to crowdsourcing workers and crowdsourcing tasks; Construct optimization problems for multi-party scenarios based on the acquired attribute sets, including optimization problems for constructing platform profits; Establish an optimization problem for maximizing worker benefits; Establish an optimization problem for maximizing task quality robustness; Establish an optimization problem of maximizing the benefit of the requester; The improved RVEA optimization algorithm is used to optimize the optimization problem of multi-party scenarios and obtain the optimization results; The obtaining of attribute sets corresponding to crowdsourcing workers and crowdsourcing tasks includes collecting attribute sets corresponding to crowdsourcing workers and crowdsourcing tasks, and extracting social attributes of workers to obtain social attributes of crowdsourcing workers in each spatiotemporal region; considering the location and equipment requirement attributes of tasks at the same time; saving the real-time attributes of workers and tasks, and generalizing the privacy location attributes of workers; The optimization problem of constructing the platform profit includes establishing the optimization problem according to the social attributes of the crowdsourcing workers, the social attributes of the crowdsourcing tasks themselves, and the profit of the platform; taking the platform profit as the optimization problem of the platform, and considering the social attributes of the workers and the tasks at the same time, the objective function of the optimization problem is expressed as: There are N workers and M tasks. The profit of the platform is represented by P, and the task allocation is represented by a binary variable X ij , where X ij = 1 if worker i is assigned to task j, otherwise 0; P ij is the payoff of worker i for performing task j; The optimization problem of maximizing worker benefits is established, including considering the benefits of each crowdsourcing worker performing tasks, the degree of adaptability of the social attributes of crowdsourcing workers to crowdsourcing tasks, and the profit of the platform. Suppose there are N workers, M tasks, the benefit of worker i performing task j is CW, the cost of the worker is COST, and the task allocation is represented by a binary variable X ij , where X ij = 1 means worker i is assigned to task j, otherwise it is 0; the objective function of the optimization problem is expressed as: The optimization problem of maximizing the robustness of task quality is established, including considering the quality of each crowdsourcing task, the adaptability of crowdsourcing workers to crowdsourcing tasks, and the evaluation of workers by the quality of tasks. There are M tasks, and the goal is to find a task allocation scheme that maximizes the robustness of task quality and makes the quality fluctuation of all tasks small. The objective function of the optimization problem is expressed as: Where Qj is the average quality of task j, is the average quality of all tasks; The optimization problem of establishing the requester's benefit maximization includes considering the quality of each crowdsourcing task, the completion time of the task, the evaluation of the crowdsourcing workers, and the importance of the requester to the task; The improved RVEA optimization algorithm is used to optimize the optimization problem of the multi-party scenario, including calculating the Pareto optimal solution for the objective function of the multi-party scenario based on the RVEA multi-objective optimization algorithm.

2. A high-dimensional optimization task allocation system considering three-party multi-objectives, executing the method as claimed in claim 1, characterized in that: include: The data acquisition module is configured to obtain a set of attributes corresponding to crowdsourcing workers and crowdsourcing tasks; The optimization problem building module is configured to build optimization problems of multiple scenarios based on the acquired attribute set, including building an optimization problem of platform profit; building an optimization problem of maximizing worker benefits; and building an optimization problem of maximizing task quality robustness. Establish an optimization problem of maximizing the benefit of the requester; The optimization module is configured to optimize the optimization problem of the multi-party scenario using the improved RVEA optimization algorithm to obtain the optimization result.

3. 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 a high-dimensional optimization task allocation method considering tripartite multi-objectives as described in claim 1.

4. 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 a high-dimensional optimization task allocation method considering tripartite multi-objectives as described in claim 1.

Images