CN117669919A - Multi-task distribution method, system, equipment and medium - Google Patents

Abstract

The application relates to the technical field of power warehouse task allocation, in particular to a multi-task allocation method, a system, equipment and a medium, and the technical scheme is as follows: acquiring a task to be allocated, selecting a historical capability index data set of similar tasks of workers according to the task to be allocated, and determining a role requirement vector of the task to be allocated; acquiring index weights, and carrying out weighted summation on the historical capability index data sets according to the index weights to obtain a comprehensive capability evaluation set; converting the comprehensive capacity evaluation set into a comprehensive capacity cloud model to obtain a comprehensive capacity moment array; and solving an allocation matrix according to the role requirement vector and the comprehensive capacity matrix, so that the capacity of workers is more comprehensively evaluated, and the efficiency and quality of task completion are improved.

Description

A multi-task allocation method, system, equipment and medium

Technical field

This application relates to the technical field of power warehouse task allocation, and in particular to a multi-task allocation method, system, equipment and medium.

Background technique

Power warehouses are an important part of the power industry, carrying key tasks such as storage, maintenance and distribution of power equipment. In order to ensure the reliability and efficiency of power supply, task distribution within the warehouse is crucial. However, power warehouse tasks typically cover a variety of jobs of varying nature and complexity, such as equipment maintenance, inventory management, logistics coordination, and more. These tasks require workers with different skills and ability levels to perform, so an intelligent task allocation method is needed to match workers' abilities and task requirements to improve work quality and efficiency.

The traditional empirical scheduling method is often inefficient and of poor quality. Therefore, in order to obtain an efficient task assignment solution, it is necessary to rely on existing optimization methods. In recent years, crowdsourcing platforms have developed rapidly. Many companies publish their work tasks to crowdsourcing platforms, solicit the best possible solutions, and improve the efficiency of task completion through distributed collaboration. Aiming at the lack of measurement of workers' ability uncertainty in crowdsourcing tasks, a software crowdsourcing task allocation method is proposed that supports fuzzy measurement of workers' abilities and role collaboration. This method uses fuzzy interval numbers to evaluate workers' multi-attribute ability matching degree, using The fuzzy analytic hierarchy process is used to calculate the comprehensive competency of workers. Task allocation in power warehouses is a general multi-task allocation problem, that is, each employee is qualified to perform these tasks, but the quality of evaluation based on the completion of these tasks by different employees varies. Algorithms based on crowdsourcing are not suitable for conventional multi-task allocation algorithms. In addition, existing research still lacks a comprehensive analysis of workers' abilities and does not comprehensively consider the impact of the matching degree of crowdsourcing tasks and workers' abilities on the allocation results. The above problems need to be solved.

Contents of the invention

In order to more comprehensively evaluate workers' abilities and improve the efficiency and quality of task completion, this application provides a multi-task allocation method, system, equipment and media, using the following technical solutions:

In the first aspect, this application provides a multi-task allocation method, including:

Obtain the tasks to be assigned, select the historical capability indicator data set of similar tasks of workers based on the tasks to be assigned, and determine the role requirement vector of the tasks to be assigned;

Obtain the indicator weight, and weight and sum the historical capability indicator data sets according to the indicator weight to obtain the comprehensive capability evaluation set;

Convert the comprehensive capability evaluation set into a comprehensive capability cloud model to obtain a comprehensive capability matrix;

The distribution matrix is obtained based on the role requirement vector and the comprehensive ability matrix solution.

Preferably, the indicators of the comprehensive ability evaluation set include work skills, work quality, work efficiency and work attitude.

Preferably, the role requirement vector represents the minimum number of tasks that the worker needs to complete.

Preferably, the comprehensive capability matrix is the matching capability of each worker for each task.

Preferably, the specific steps for converting the comprehensive capability evaluation set into a comprehensive capability cloud model to obtain the comprehensive capability matrix are:

Introducing cloud model theory, cm={Ex,En,He},

Among them, QS is the comprehensive ability evaluation set; Ex is the average value of the set QS; σ is the standard deviation of Ex; S ² is the sample variance of Ex; N is the number of samples in the set QS. Euclidean distance is used to calculate the similarity. The formula is as follows:

Convert the worker's comprehensive ability evaluation set into a comprehensive ability cloud model. When the worker has only completed the task once in the historical record, p is the comprehensive ability value, and the cloud model is {p,0,0}; for m workers The cloud model of comprehensive capability evaluation is described as:

in, It is a cloud model of workers’ comprehensive capabilities for tasks;

Define cm ^- and cm ⁺ as the worst and best states of worker ability, and their expressions are as follows:

From this, we can get the worker’s competency for the task as:

Obtain the comprehensive capability matrix Q.

Preferably, it also includes:

Based on the analysis of the comprehensive capability matrix and the allocation matrix, the work group performance indicating the quality of completion of all tasks is obtained.

Preferably, the historical capability indicator data set represents several historical capability indicator values in which a worker completed the task to be assigned.

In the second aspect, this application provides a multi-task allocation system, including:

The first acquisition module: used to obtain the tasks to be assigned, select the historical capability indicator data set of similar tasks of workers based on the tasks to be assigned, and determine the role requirement vector of the tasks to be assigned;

The second acquisition module: used to obtain indicator weights, and weighted summation of historical capability indicator data sets based on the indicator weights to obtain a comprehensive capability evaluation set;

Capability evaluation module: used to convert the comprehensive capability evaluation set into a comprehensive capability cloud model to obtain a comprehensive capability matrix;

Task allocation module: used to obtain the allocation matrix based on the role requirement vector and the comprehensive capability matrix solution.

In a third aspect, the present application provides a multi-task allocation device, including a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to perform the multi-task allocation method as described above. .

In a fourth aspect, the present application provides a computer-readable storage medium in which a computer program is stored, wherein the computer program is configured to execute the multi-task allocation method as described above when running.

To sum up, compared with the existing technology, the beneficial effects brought by the technical solution provided by this application at least include:

This application selects the historical capability indicator data set of similar tasks of workers based on the tasks to be assigned, and determines the role requirement vector of the assigned tasks. After obtaining the historical capability indicator data set of workers, the comprehensive capability evaluation is obtained based on the weighted sum of the weights of each selected indicator. Collection, the workers' historical comprehensive ability evaluation values are converted into a comprehensive ability cloud model, and after further processing, a comprehensive ability matrix is obtained to more comprehensively and accurately evaluate the workers' abilities, and based on this, an allocation matrix is obtained to allocate tasks flexibly and accurately. , improve the efficiency and quality of task completion.

Description of drawings

Figure 1 is a schematic flowchart of a multi-task allocation method according to an embodiment of the present application.

Figure 2 is the ability evaluation index information of worker a ₄ described in the embodiment of this application.

Figure 3 is a historical comprehensive ability set of worker a ₄ according to the embodiment of this application.

Figure 4 is a comprehensive capability cloud model of worker a ₄ for each sub-task in the embodiment of this application.

Figure 5 is an employee task quality evaluation form according to the embodiment of this application.

Figure 6 is a schematic module diagram of a multi-task allocation system according to the embodiment of the present application.

Explanation of reference symbols:

1. The first acquisition module; 2. The second acquisition module; 3. Capability evaluation module; 4. Task allocation module.

Detailed ways

The present application will be further described in detail below with reference to FIGS. 1-6 . The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to be limiting.

Referring to Figure 1, a multi-task allocation method involved in this application specifically includes:

Step S1: Obtain the tasks to be assigned, select the historical capability indicator data set of similar tasks of workers based on the tasks to be assigned, and determine the role requirement vector of the tasks to be assigned;

Step S2: Obtain the indicator weights, and weight and sum the historical capability indicator data sets according to the indicator weights to obtain the comprehensive capability evaluation set;

Step S3: Convert the comprehensive capability evaluation set into a comprehensive capability cloud model to obtain a comprehensive capability matrix;

Step S4: Obtain the distribution matrix based on the role requirement vector and comprehensive ability matrix solution.

Specifically, the embodiment of the present application combines cloud model theory with the E-CARGO algorithm to solve the problem of power warehouse task allocation. Select the historical capability indicator data set of similar tasks of workers based on the tasks to be assigned, and determine the role requirement vector of the assigned tasks. After obtaining the historical capability indicator data set of workers, the comprehensive capability evaluation set is obtained based on the weighted sum of the weights of each selected indicator. Through cloud model theory, workers' abilities can be comprehensively and accurately assessed, taking into account various uncertain factors such as skill level, experience, and adaptability. The worker's historical comprehensive ability evaluation value is converted into a comprehensive ability cloud model. After further processing, a comprehensive ability matrix is obtained to evaluate the worker's ability more comprehensively and accurately. The application of the E-CARGO algorithm will make task allocation more adaptive and intelligent. , flexibly adjust according to the nature of the task and the ability of the workers, and use this as a basis to obtain the allocation matrix, allocate tasks flexibly and accurately, and improve the efficiency and quality of task completion.

In conventional multi-tasking, each worker is qualified for the tasks in the power warehouse, but in fact the evaluation quality of different workers completing these tasks varies. The tasks to be assigned are task sets Ω={Ω ₁ ,Ω _n ,...,Ω _n }, where Ω _n is the nth task, and each task is completed by a worker. Obtain the historical evaluation information of workers from managers, let Λ = {Λ ₁ , Λ ₂ ,..., Λ _m }, Λ _m represents the m-th worker, and each worker can complete multiple tasks. make Represents the k-th historical capability index value of worker Λ _m completing task Ω _n . The historical capability indicator data set represents several historical capability indicator values of one of the workers completing the task to be assigned. There are differences in the ratings of workers completing the same task, and the abilities of different workers are different, and the quality of completing the same task is also different.

The calculation of the worker's comprehensive ability matrix Q is mainly based on the worker's historical evaluation. The comprehensive ability matrix is the matching ability of each worker for each task, and is then obtained based on the cloud theory model. The historical evaluation of workers mainly includes four parts, namely work skills, work quality, work efficiency, and work attitude. The range of the four indicators is [0,1], accounting for 2:5:1:2 respectively. This ratio comes from the warehouse management side. In this way, we can obtain a reasonable comprehensive ability evaluation set QS of workers for a specific task.

As one of the implementation methods, the comprehensive capability evaluation set is converted into a comprehensive capability cloud model, and the specific steps to obtain the comprehensive capability matrix are:

Introducing the cloud model theory, cm = {Ex, En, He}, where Ex is expectation, En is entropy, and He is super entropy. Expectation is the most representative value of comprehensive ability, entropy represents the granularity range of comprehensive ability, and super entropy represents It is the uncertainty of the granularity of comprehensive capabilities. The three numerical characteristics of the cloud model can be calculated by the following formula:

Among them, QS is the comprehensive ability evaluation set; Ex is the average value of the set QS; σ is the standard deviation of Ex; S ² is the sample variance of Ex; N is the number of samples of the set QS. Different tasks require different skills. It is important to identify the differences between different workers on different tasks by computing similarities between cloud models. The embodiment of this application uses Euclidean distance to calculate similarity, and the calculation formula is as follows:

Convert the worker's comprehensive ability evaluation set into a comprehensive ability cloud model. When the worker has only completed the task once in the historical record, p is the only comprehensive ability value, and the cloud model is {p,0,0}; for m The cloud model of workers’ comprehensive ability evaluation is described as:

in, It is a cloud model of workers' comprehensive ability for tasks; workers have fluctuations and will have different comprehensive ability values when completing tasks. The higher the Ex value, the stronger the worker's comprehensive ability, the lower the En and He values, the more stable the worker's comprehensive ability. Therefore, cm ^- and cm ⁺ are defined as the worst and best states of the worker's ability, and their expressions are as follows :

From this we can get that worker ai’s competency for task rj can be calculated using the following formula:

Obtain the comprehensive capability matrix Q. The larger the value of Q[i,j], the stronger the comprehensive ability of worker a _i for task r _j .

After obtaining the evaluation matrix Q, the distribution matrix needs to be solved. Since the maximum value in Q[i,j] is 1, in order to convert the maximum value of the allocation problem into the minimum value problem, here, the matrix C is defined, where C[i,j]=1-Q[ i, j]. At this time, the allocation problem of the Q matrix is transformed into the allocation of the C matrix.

The solution process used in the embodiments of this application is a conversion method for generalized assignment problems. Here, α _i is defined as the minimum number of tasks performed by person i, α′ _i is the maximum number of tasks performed by person i; β _j is defined as the minimum number of people performing task j, and β′ _j is the maximum number of people performing task j; d is the total number of assigned persons. The specific conversion process is as follows:

Copy each row i of the C matrix by α′ _i -1 rows, and then copy j of the copied matrix by β′ _j -1 to obtain a new matrix:

This matrix has a total of OK,/> Column, C ^ij (i=1,…,m; j=1,…,n) is a subarray with α′ _i rows and β′ _j columns, and each element is C _ij .

set up y=max(t,v,d), p=uy, q=sy, right/> Add p rows and q columns to get an extended matrix:

Among them, G ^kj is a matrix with 1 column and β′ _j column, k=1,2,...,p; j=1,2,...,n.

here in

Here F ^il is a matrix with α′ _i rows and 1 column, i=1,2,…,m; l=1,2,…,q, Among them/>

Here's Where M is a large enough number,/> is a sufficiently large number larger than M. Then the solution to matrix B is the solution to matrix Q. The matrix B here is solved using the Hungarian algorithm.

After obtaining the allocation matrix based on the role requirement vector and the comprehensive capability matrix solution, the E-CARGO model's task allocation problem for workers can be abstracted as: ∑∷=<E,C,O,R,A,G>. Among them: E represents a problem environment involving multiple workers and multiple tasks; C is a class collection of abstract concepts in E; O is a specific object collection related to C; R is a collection of tasks to be assigned, represented by role; A is a candidate worker set, represented by agent; G is a work group group, that is, a worker team established by a task allocation algorithm. In the embodiment of this application, tasks are mapped to roles, and the task set Ω={Ω ₁ ,Ω ₂ ,...,Ω _n } corresponds to the role set R={r ₁ ,r ₂ ,...,r _n } ; Mapping workers as agents, the worker set Λ = {Λ ₁ , Λ ₂ ,..., Λ _m } corresponds to the agent set A = {a ₁ , a ₂ ,..., a _m }, and the worker's task completion is Agents play roles, a collection of indicators of worker capabilities The corresponding agent qualification set is In addition, it is necessary to make a simplified definition of the concepts related to this issue in E-CARGO.

The worker’s task demand vector L. L _j represents the minimum number of tasks that worker Λ _j needs to complete. If L _j ≥ 1, it means that Λ _j needs to complete multiple tasks.

Worker comprehensive ability evaluation matrix Q. Q is an m×n matrix, Q[i,j]∈[0,1](0≤i≤m; 0≤j≤n). It represents the degree of competence of worker a _i for the task r _j to be assigned. The comprehensive ability of workers can be measured based on factors such as work skills, work quality, work efficiency, and work attitude.

Role assignment matrix X. X is an m×n matrix, where X[i,j]∈{0,1} (0≤i≤m; 0≤j≤n). If Assigned to role j, that is, worker i is assigned to role j. The agent at this time is called an assigned agent. If X[i,j]=0, it means that agent i is not assigned to role j.

As one of the implementation methods, it also includes:

Based on the analysis of the comprehensive capability matrix and the allocation matrix, the work group performance indicating the quality of completion of all tasks is obtained.

Specifically, work group performance ρ. All workers assigned tasks form a work group G. ρ represents the sum of ability values of all workers in G. The larger ρ is, the higher the completion quality of all tasks is. For a set of tasks, it is necessary to maximize the worker's ability and ensure that the ρ of all tasks is the highest. A given task is not necessarily assigned to the most competent person.

Based on the above definition, the objective function of the multi-task allocation problem in this project is:

s.t X[i,j]∈{0,1}; 0≤i<m,0≤j<n;

As one of the implementation methods, there are currently a task set {r ₁ , r ₂ , r ₃ , r ₄ , r ₅ , r ₆ } and a worker set {a ₁ , a ₂ , a ₃ , a ₄ }. That is, there are 6 tasks that require 4 workers to complete. Today every employee is competent in these tasks, but the quality of assessments based on how well each employee completes these tasks varies. The role requirement vector is L=[1,1,2,2]. Worker a ₁ completes one task, worker a ₂ completes one task, and workers a ₃ and a ₄ each complete two tasks. The indicator set is k={k ₁ , k ₂ , k ₃ , k ₄ }, which respectively represent work skills, work quality, work efficiency, and work attitude. The following uses the historical ability index information of worker a ₄ in completing similar tasks as an example to calculate its evaluation value on each task. Refer to Figure 2 for worker _a4 ’s ability evaluation index information.

According to the set weight of each indicator 2:5:1:2, the historical comprehensive ability set of worker a ₄ and the comprehensive ability cloud model of worker a ₄ for each sub-task can be obtained. Refer to Figure 3 for the historical comprehensive ability set of worker a ₄ .

The comprehensive capability cloud model of worker a ₄ for each sub-task is shown in Figure 4.

After calculating the comprehensive capability cloud model of all workers for each sub-task, cm ^- ={0.37,0.1,0.03}, cm ⁺ ={0.857,0,0} are obtained. From this, we can get the evaluation matrix Q of the comprehensive ability of the worker set {a ₁ , a ₂ , a ₃ , a ₄ } to the task set {r ₁ , r ₂ , r ₃ , r ₄ , r ₅ , r ₆ }, specifically The correspondence is shown in Figure 5, which is the employee task quality evaluation form.

From the above Q matrix, according to C[i,j]=1-Q[i,j], the C matrix can be obtained:

Here L = [1, 1, 2, 2], which is equivalent to one and only one person completing each task. A ₁ and a ₂ each complete one task, and a ₃ and a ₄ each complete two tasks. Corresponding vector α=[1,1,2,2], α′=[1,1,2,2], β=[1,1,1,1,1,1], β′=[1,1 ,1,1,1,1], corresponding to this The total number of people on the task is d=4, y=max(t,v,d)=6, p=uy=0, q=sy=0. From the above calculation, it can be concluded that the matrix/> It is the same matrix as matrix B. Copying each row i of the C matrix to the α′ _i -1 row means copying one row each of the third and fourth rows of the C matrix; since β′ _j -1=0, the column direction of the C matrix does not change. Finally, we get the following /> matrix:

This matrix is solved using the Hungarian algorithm:

According to the above solution, the final assigned matrix X can be obtained:

_{From the above allocation matrix ​ ​ ​ ​ 5} . The maximum ρ is 5.404.

Referring to Figure 5, a multi-task allocation system is provided for this embodiment of the present application. The system includes:

The first acquisition module 1: used to obtain the tasks to be assigned, select the historical capability indicator data set of similar tasks of workers based on the tasks to be assigned, and determine the role requirement vector of the tasks to be assigned;

Second acquisition module 2: used to obtain indicator weights, and weighted summation of historical capability indicator data sets based on the indicator weights to obtain a comprehensive capability evaluation set;

Capability evaluation module 3: used to convert the comprehensive capability evaluation set into a comprehensive capability cloud model to obtain a comprehensive capability matrix;

Task allocation module 4: used to obtain the allocation matrix based on the role requirement vector and comprehensive capability matrix solution.

An embodiment of the present application provides a multi-task allocation device, including a memory and a processor. The memory stores a computer program, and the processor is configured to run the computer program to execute the multi-task allocation method as described above.

Embodiments of the present application provide a computer-readable storage medium in which a computer program is stored, wherein the computer program is configured to execute the multi-task allocation method as described above when running.

Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the devices and products described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be described again here.

In the several embodiments provided in this application, it should be understood that the disclosed methods, systems, devices and program products can be implemented in other ways.

In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or software functional units.

As mentioned above, the above embodiments are only used to illustrate the technical solution of the present application, but not to limit it. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still make the foregoing technical solutions. The technical solutions described in each embodiment may be modified, or some of the technical features may be equivalently replaced; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions in each embodiment of the present application.

Claims (10)

1. A method of multitasking, comprising:

acquiring a task to be allocated, selecting a historical capability index data set of similar tasks of workers according to the task to be allocated, and determining a role requirement vector of the task to be allocated;

acquiring index weights, and carrying out weighted summation on the historical capability index data sets according to the index weights to obtain a comprehensive capability evaluation set;

converting the comprehensive capacity evaluation set into a comprehensive capacity cloud model to obtain a comprehensive capacity moment array;

and solving an allocation matrix according to the character requirement vector and the comprehensive capacity matrix.

2. The method of claim 1, wherein the indicators of the aggregate competence assessment set include work skills, work quality, work efficiency, and work attitude.

3. The multitasking method of claim 1, wherein said character requirement vector represents a minimum number of tasks a worker needs to accomplish.

4. The method of claim 1, wherein the aggregate capacity matrix is a match capacity of each worker for each task.

5. The method for assigning multiple tasks according to claim 4, wherein the step of converting the comprehensive ability evaluation set into a comprehensive ability cloud model to obtain a comprehensive ability matrix comprises the steps of:

introducing a cloud model theory, cm= { Ex, en, he },

wherein QS is a comprehensive capability evaluation set; ex is the average of the set QS; sigma is the standard deviation of Ex; s is S ² Is the sample variance of Ex; n is the number of samples of the set QS, and the Euclidean distance is used for calculating the similarity, and the calculation formula is as follows:

converting the comprehensive capability evaluation set of the workers into a comprehensive capability cloud model, wherein p is a comprehensive capability value and the cloud model is { p, 0}, under the condition that the workers only finish one task in the history record; the cloud model for comprehensive ability evaluation of m workers is described as:

wherein,the comprehensive capacity cloud model of the worker for the task;

definition cm ^- And cm ⁺ The expression is as follows, which is the worst and best state of worker ability:

the competence of the worker for the task can be obtained by the method that:

and obtaining the comprehensive energy moment array Q.

6. The method of multitasking assignment of claim 1, further comprising:

and analyzing according to the comprehensive capacity matrix and the distribution matrix to obtain the performance of the workgroup for indicating the completion quality of all tasks.

7. The method of multitasking in accordance with claim 1, characterized in that said historical capability index data set represents a number of historical capability index values for one of the workers to complete a task to be assigned.

8. A multi-tasking distribution system comprising:

a first acquisition module: the method comprises the steps of acquiring a task to be allocated, selecting a historical capability index data set of similar tasks of workers according to the task to be allocated, and determining a role requirement vector of the task to be allocated;

and a second acquisition module: the comprehensive capacity evaluation method comprises the steps of obtaining index weights, and carrying out weighted summation on a historical capacity index data set according to the index weights to obtain a comprehensive capacity evaluation set;

capability evaluation module: the comprehensive capacity evaluation set is used for converting the comprehensive capacity evaluation set into a comprehensive capacity cloud model to obtain a comprehensive capacity moment array;

the task allocation module: for solving an allocation matrix based on the character requirement vector and the comprehensive capacity matrix.

9. A multitasking apparatus comprising a memory storing a computer program and a processor arranged to run the computer program to perform a multitasking method as claimed in any of claims 1-7.

10. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein a computer program, wherein the computer program is arranged to execute the multitasking method of any of claims 1-7 when run.