CN117436627A - Task allocation method, device, terminal equipment and medium - Google Patents
Task allocation method, device, terminal equipment and medium Download PDFInfo
- Publication number
- CN117436627A CN117436627A CN202311085576.8A CN202311085576A CN117436627A CN 117436627 A CN117436627 A CN 117436627A CN 202311085576 A CN202311085576 A CN 202311085576A CN 117436627 A CN117436627 A CN 117436627A
- Authority
- CN
- China
- Prior art keywords
- task
- executed
- tasks
- allocation
- group
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000013528 artificial neural network Methods 0.000 claims abstract description 59
- 239000011159 matrix material Substances 0.000 claims abstract description 58
- 238000004422 calculation algorithm Methods 0.000 claims description 37
- 230000002068 genetic effect Effects 0.000 claims description 28
- 230000006870 function Effects 0.000 claims description 22
- 239000013598 vector Substances 0.000 claims description 21
- 230000008569 process Effects 0.000 claims description 9
- 238000013507 mapping Methods 0.000 claims description 8
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 10
- 238000011156 evaluation Methods 0.000 description 8
- 238000005457 optimization Methods 0.000 description 8
- 238000011161 development Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 230000035772 mutation Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 239000010410 layer Substances 0.000 description 3
- 238000013468 resource allocation Methods 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 210000000349 chromosome Anatomy 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 239000002355 dual-layer Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06311—Scheduling, planning or task assignment for a person or group
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Marketing (AREA)
- Bioinformatics & Computational Biology (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Evolutionary Biology (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Development Economics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Genetics & Genomics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a task allocation method, a device, terminal equipment and a computer readable storage medium, wherein the method comprises the following steps: determining a plurality of tasks to be executed and a plurality of task execution groups; acquiring task feature sequences of the tasks to be executed and task subgroup sequences of the task execution subgroups, and determining task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network; determining task dependency relations among the plurality of tasks to be executed, and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed; and acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed. The invention can realize the accurate allocation of the tasks, thereby improving the task processing efficiency.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a task allocation method, a task allocation device, a terminal device, and a computer readable storage medium.
Background
With the advancement of digital transformation, software development projects are increasingly large in scale. In this context, software development teams face a series of challenges in resource configuration and task scheduling.
For example, with existing resource allocation and task scheduling methods, tasks and required resources are commonly allocated to various software development teams through genetic algorithms.
However, in such task allocation and task scheduling methods, many key factors, such as actual performance of executing tasks and resources required for executing tasks, are not considered, so that accurate task allocation and resource scheduling cannot be performed for each development team, resulting in low processing efficiency of subsequent tasks.
Disclosure of Invention
The invention mainly aims to provide a task allocation method, a device, terminal equipment and a computer readable storage medium, aiming at realizing accurate task allocation and further improving task processing efficiency.
In order to achieve the above object, the present invention provides a task allocation method, which includes the steps of:
determining a plurality of tasks to be executed and a plurality of task execution groups;
acquiring task feature sequences of the tasks to be executed and task subgroup sequences of the task execution subgroups, and determining task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network;
Determining task dependency relations among the plurality of tasks to be executed, and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed;
and acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed.
Optionally, before the step of obtaining the task feature sequences of the tasks to be executed and the task group sequences of the task execution groups, and determining the task matching weights between the task group sequences and the task feature sequences through a preset neural network, the method further includes:
acquiring task weights of the tasks to be executed aiming at each task to be executed in a current time window;
constructing the preset neural network according to the sequence sizes of the task feature sequence and the task group sequence;
the step of determining the task matching weight between the task group sequence and the task feature sequence through a preset neural network comprises the following steps:
and mapping the task feature sequence and the task group sequence into equal-length Boolean vectors respectively, and taking the Boolean vectors as the input of the preset neural network to obtain task matching weights between the task group sequence and the task feature sequence output by the preset neural network, wherein the significance probability between the task matching weights and the task weights of the tasks to be executed is smaller than a preset difference threshold.
Optionally, after the step of mapping the task feature sequence into boolean vectors with equal length, and taking the boolean vectors as the input of the preset neural network, obtaining a task matching weight between the task execution subgroup output by the preset neural network and the task feature sequence of the task to be executed, the method further includes:
if the significance probability between the task matching weight and the task weight is larger than the preset difference threshold, training the preset neural network in a back propagation mode until the significance probability between the task matching weight and the task weight is smaller than the preset difference threshold.
Optionally, the step of determining task dependencies between the plurality of tasks to be performed includes:
acquiring the dependency relationship among the plurality of tasks to be executed;
and constructing a task dependency graph according to the dependency relationship, wherein each node of the task dependency graph represents each task to be executed, and the directed edge of the task dependency graph represents the dependency relationship.
Optionally, the step of constructing a resource requirement constraint matrix when the task execution group executes the task to be executed includes:
Acquiring the resource demand of the task execution group to execute the task to be executed;
and constructing a resource demand constraint matrix according to the resource demand, wherein the resource demand constraint matrix comprises the resource demand of each task to be executed for executing each task of each task execution group.
Optionally, before the step of obtaining the optimal task allocation scheme based on the task dependency relationship, the resource requirement constraint matrix and the task matching weight and in combination with the task time limit corresponding to each task to be executed, the method further includes:
constructing an initialized population, wherein each individual in the initialized population characterizes a task allocation scheme;
acquiring dependency relation weight according to the task dependency graph;
acquiring inter-group resource load balancing weights corresponding to the task execution groups according to the resource demand constraint matrix;
and carrying out weight superposition on the task matching weight, the dependency relationship weight and the inter-group resource load balancing weight to obtain an adaptability function, so as to obtain a task allocation and scheduling optimal scheme by combining task execution group corresponding task time limits according to the adaptability function and through a genetic algorithm.
Optionally, the step of obtaining the optimal task allocation scheme based on the task dependency relationship, the resource requirement constraint matrix and the task matching weight and in combination with the task time limit corresponding to each task to be executed includes:
determining a task set to be executed corresponding to each task execution group;
acquiring task time limits of all the tasks to be executed in the task set to be executed, wherein the task time limits comprise task execution time and task execution deadline;
calculating the urgent value of each task to be executed according to the task execution time and the task execution deadline and by combining the current time;
and optimizing the task allocation scheme through a genetic algorithm according to the fitness function, and sequencing a plurality of tasks to be executed according to the urgent value in the optimizing process until the task allocation optimal scheme is obtained.
In order to achieve the above object, the present invention also provides a task allocation device, including:
the first determining module is used for determining a plurality of tasks to be executed and a plurality of task execution groups;
the second determining module is used for acquiring task feature sequences of the plurality of tasks to be executed and task subgroup sequences of the plurality of task execution subgroups, and determining task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network;
The third determining module is used for determining task dependency relations among the plurality of tasks to be executed and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed;
the acquisition module is used for acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed.
In order to achieve the above object, the present invention also provides a terminal device including a memory, a processor, and a task allocation program stored on the memory and executable on the processor, which when executed by the processor, implements the steps of the task allocation method as described above.
In addition, in order to achieve the above object, the present invention also proposes a computer-readable storage medium having stored thereon a task allocation program which, when executed by a processor, implements the steps of the task allocation method as described above.
To achieve the above object, the present invention also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the task allocation method as described above.
The invention provides a task allocation method, a device, a terminal device, a computer readable storage medium and a computer program product, which are used for determining a plurality of tasks to be executed and a plurality of task execution groups; acquiring task feature sequences of the tasks to be executed and task subgroup sequences of the task execution subgroups, and determining task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network; determining task dependency relations among the plurality of tasks to be executed, and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed; and acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed.
Compared with the task allocation mode in the prior art, the task allocation optimal scheme can be solved through a preset neural network, the task matching weight between the task group sequence and the task feature sequence, the task dependency relationship among a plurality of tasks to be executed, the resource demand constraint matrix when each task execution group executes the task to be executed and the task execution time corresponding to the task to be executed. Therefore, the invention considers the constraint relation between the task execution group and the task to be executed, the constraint among the tasks to be executed, the constraint of resources required by the task to be executed and the constraint of the task execution time, and determines the optimal task allocation scheme through the genetic algorithm under a plurality of constraint conditions, so that the optimal task allocation scheme can accurately adapt to the task execution capacity and the resource requirement of each group, the task accurate allocation is realized, and the subsequent task processing efficiency is further improved.
Drawings
FIG. 1 is a schematic diagram of a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a task allocation flow according to an embodiment of the task allocation method of the present invention;
FIG. 3 is a first flow chart illustrating an embodiment of a task allocation method according to the present invention;
FIG. 4 is a second flow chart of an embodiment of a task allocation method according to the present invention;
FIG. 5 is a functional block diagram of an embodiment of a task assigning device according to the present invention.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, fig. 1 is a schematic device structure of a hardware running environment according to an embodiment of the present invention.
The terminal equipment in the embodiment of the invention can be a mobile phone, a tablet personal computer, a server or other network equipment and the like, and can be used for realizing efficient and accurate task allocation.
As shown in fig. 1, the terminal device may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
Those skilled in the art will appreciate that the device structure shown in fig. 1 is not limiting of the task distribution device and may include more or fewer components than shown, or may combine certain components, or may be a different arrangement of components.
As shown in fig. 1, an operation, a network communication module, a user interface module, and a task allocation program may be included in the memory 1005 as one type of computer storage medium. Operations are programs that manage and control device hardware and software resources, support task allocation programs, and other software or program runs. In the device shown in fig. 1, the user interface 1003 is mainly used for data communication with the client; the network interface 1004 is mainly used for establishing communication connection with a server; and the processor 1001 may be configured to call a task allocation program stored in the memory 1005 and perform the following operations:
determining a plurality of tasks to be executed and a plurality of task execution groups;
acquiring task feature sequences of the tasks to be executed and task subgroup sequences of the task execution subgroups, and determining task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network;
Determining task dependency relations among the plurality of tasks to be executed, and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed;
and acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed.
Further, before the step of obtaining the task feature sequences of the tasks to be executed and the task group sequences of the task execution groups, and determining the task matching weights between the task group sequences and the task feature sequences through the preset neural network, the processor 1001 may be configured to call a task allocation program stored in the memory 1005, and perform the following operations:
acquiring task weights of the tasks to be executed aiming at each task to be executed in a current time window;
constructing the preset neural network according to the sequence sizes of the task feature sequence and the task group sequence;
the processor 1001 may be configured to call a task allocation program stored in the memory 1005 and perform the following operations:
And mapping the task feature sequence and the task group sequence into equal-length Boolean vectors respectively, and taking the Boolean vectors as the input of the preset neural network to obtain task matching weights between the task group sequence and the task feature sequence output by the preset neural network, wherein the significance probability between the task matching weights and the task weights of the tasks to be executed is smaller than a preset difference threshold.
Further, after the step of mapping the task feature sequences into boolean vectors with equal length and taking the boolean vectors as the input of the preset neural network to obtain the task matching weights between the task execution subgroup output by the preset neural network and the task feature sequences of the tasks to be executed, the processor 1001 may be configured to invoke a task allocation program stored in the memory 1005 and perform the following operations:
if the significance probability between the task matching weight and the task weight is larger than the preset difference threshold, training the preset neural network in a back propagation mode until the significance probability between the task matching weight and the task weight is smaller than the preset difference threshold.
Further, the processor 1001 may be configured to call a task allocation program stored in the memory 1005, and perform the following operations:
acquiring the dependency relationship among the plurality of tasks to be executed;
and constructing a task dependency graph according to the dependency relationship, wherein each node of the task dependency graph represents each task to be executed, and the directed edge of the task dependency graph represents the dependency relationship.
Further, the processor 1001 may be configured to call a task allocation program stored in the memory 1005, and perform the following operations:
acquiring the resource demand of the task execution group to execute the task to be executed;
and constructing a resource demand constraint matrix according to the resource demand, wherein the resource demand constraint matrix comprises the resource demand of each task to be executed for executing each task of each task execution group.
Further, before the step of obtaining the optimal task allocation scheme based on the task dependency relationship, the resource requirement constraint matrix, and the task matching weight and in combination with the task time limit corresponding to each task to be executed, the processor 1001 may be configured to call a task allocation program stored in the memory 1005, and perform the following operations:
Constructing an initialized population, wherein each individual in the initialized population characterizes a task allocation scheme;
acquiring dependency relation weight according to the task dependency graph;
acquiring inter-group resource load balancing weights corresponding to the task execution groups according to the resource demand constraint matrix;
and carrying out weight superposition on the task matching weight, the dependency relationship weight and the inter-group resource load balancing weight to obtain an adaptability function, so as to obtain a task allocation and scheduling optimal scheme by combining task execution group corresponding task time limits according to the adaptability function and through a genetic algorithm.
Further, the processor 1001 may be configured to call a task allocation program stored in the memory 1005, and perform the following operations:
determining a task set to be executed corresponding to each task execution group;
acquiring task time limits of all the tasks to be executed in the task set to be executed, wherein the task time limits comprise task execution time and task execution deadline;
calculating the urgent value of each task to be executed according to the task execution time and the task execution deadline and by combining the current time;
And optimizing the task allocation scheme through a genetic algorithm according to the fitness function, and sequencing a plurality of tasks to be executed according to the urgent value in the optimizing process until the task allocation optimal scheme is obtained.
In light of the above background description of the application, for existing resource allocation and task scheduling, it is common to allocate tasks and required resources to each software development team through genetic algorithms, for example, as shown in fig. 2, including the following steps:
(1) Problem modeling and coding: abstracting task allocation and scheduling problems into an optimization model, designing a proper coding scheme, and representing feasible solutions in a solution space as a data structure which can be processed by a computer;
(2) Initializing a population: randomly generating a set of initial individuals to form a population, each individual representing a scheduling scheme;
(3) And (3) adaptability evaluation: setting a fitness function according to an optimization target, evaluating each individual in the population, and measuring the performance of a scheduling scheme corresponding to the individual on a problem;
(4) Termination condition: setting a condition of stopping the algorithm from two aspects of calculation time and solution precision, if the condition is not satisfied, continuing to execute, otherwise, jumping to the step VIII to output an optimal individual;
(5) Selection, crossover and mutation: selecting excellent individuals as parents according to fitness by adopting a proper selection operator, and generating new individuals through crossover and mutation operations;
(6) Updating the population: combining the newly generated individuals and part of the parent individuals to form an updated population;
(7) Iterative optimization: repeatedly executing the steps III to VI, and continuously improving the population through the evolution process;
(8) Outputting the optimal individual and decoding: after the termination condition is met, selecting an individual with the highest fitness from the final population as an optimal solution, namely a task allocation and scheduling scheme for the problem.
According to the above manner, the prior art can realize task allocation, resource scheduling and the like.
However, this approach has at least the following problems:
(1) The starting time of the new task is long: the tasks are simply classified, the similarity among the tasks is not considered, and when a new task is generated, the weight is filled only in a random initialization mode, so that the algorithm accuracy is affected;
(2) Lack of self-learning mechanisms: when the fitness is evaluated, static weight parameters are adopted, learning curves and adaptability of different development groups are not considered, so that an algorithm cannot be dynamically adjusted according to actual performances of the different development groups, and the improvement of the overall performance is limited;
(3) The algorithm convergence speed is low: aiming at the problem of multi-task and multi-resource constraint of adding time dimension, the effective solution of the genetic algorithm has large search space and slower convergence speed, and a large number of iterations are needed to obtain a better solution;
(4) Fitness evaluation direction is single: only the relevance between the task and the group is considered, and the task dependence and the resource constraint in the actual development environment are not evaluated.
In order to solve the above problems, the present invention provides a task allocation method, which aims to perform task accurate allocation for each task execution group, thereby improving task processing efficiency.
Referring to fig. 3, fig. 3 is a flowchart illustrating a task allocation method according to a first embodiment of the present invention.
Embodiments of the present invention provide embodiments of a task allocation method, it being noted that although a logical order is illustrated in the flowchart, in some cases the steps illustrated or described may be performed in a different order than that illustrated herein.
Specifically, the task allocation method in this embodiment includes the following steps:
step S10, determining a plurality of tasks to be executed and a plurality of task execution groups;
in this embodiment, for a development of a project, the terminal device may determine a plurality of tasks to be executed and a plurality of task execution groups in advance, where each task execution group includes at least one developer.
Specifically, for example:
(1) And (3) data acquisition: setting a data acquisition window with a window interval of tau and a time span of T;
(2) And (3) data processing: assuming that n task execution groups currently exist, historical performance data of m tasks are collected in a data acquisition window, and the data are cleaned, converted and normalized and then expressed as the following data matrix:
task execution time matrix E nm Wherein E is ij Representing the time it takes for the ith team to perform the jth task;
II task completion quality matrix Q nm Wherein Q is ij A quality score representing completion of the j-th task by the i-th panel;
III group resource utilization vector U n Wherein U is i Representing the resource utilization rate of the ith subgroup, i epsilon n;
vector of overdue days D m Wherein j e m represents the overdue days of the j-th task;
(3) Determining task weight W ij ＝T ij *U i *P j Wherein, the method comprises the steps of, wherein,scoring task execution efficiency, U i ＝U i Is a resource utilization factor->Is the overdue penalty factor (α is used to control the effect of overdue days on penalty factor).
Through the above operation, the terminal device confirmsA plurality of task execution groups and a plurality of tasks to be executed within a data acquisition window are defined. In addition, the present embodiment also determines the task actual weight W of each task to be executed ij Later based on the actual weight W of the task ij The neural network is optimized.
Step S20, task feature sequences of the tasks to be executed and task subgroup sequences of the task execution subgroups are obtained, and task matching weights between the task subgroup sequences and the task feature sequences are determined through a preset neural network;
in this embodiment, after acquiring a plurality of task execution subgroups and a plurality of tasks to be executed, the terminal device may acquire task feature sequences corresponding to the plurality of tasks to be executed and task subgroup sequences of the plurality of task execution subgroups, and may further determine task matching weights between the task subgroup sequences and the task feature sequences by presetting a neural network.
Specifically, for example, the terminal device may obtain task sequences T corresponding to a plurality of tasks to be performed: { T 1 ，T 2 ，T 3 ,.. } task group sequence Z: { Z 1 ，Z 2 ，Z 3 ,., further performing feature abstraction on the sequence to obtain a task feature sequence with the size of xTask group sequence of size yEach feature in the sequence is not subdivided and independent of the other.
Step S30, determining task dependency relationships among the plurality of tasks to be executed, and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed;
In this embodiment, since there may be a dependency relationship between tasks to be executed, for example, task T to be executed a Requiring a task T to be performed b Can be executed after completion, therefore, the terminal device needs to determine any of a plurality of tasks to be executedBusiness dependency relationship.
Moreover, since the resources required for executing the tasks are limited, the terminal device also needs to determine the resources required by the task execution group for executing each task to be executed in advance and construct a resource demand constraint matrix.
And step S40, acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed.
In this embodiment, after acquiring the task dependency relationship, the resource requirement constraint matrix and the task matching weight, the terminal device may acquire a task allocation optimal scheme through a genetic algorithm according to the task dependency relationship, the resource requirement constraint matrix and the task matching weight and in combination with task execution time of each task to be executed.
In this embodiment, as shown in fig. 4, before the genetic algorithm is used to perform the selection, the intersection and the mutation, the fitness function may be constructed according to the task dependency relationship, the resource requirement constraint matrix and the task matching weight output by the preset neural network, instead of the genetic algorithm shown in fig. 2. Therefore, the method and the device can evaluate the adaptability of the constraint between the group and the task, the constraint between the group and the constraint between the task, and enhance the evaluation accuracy of the scheme.
In this embodiment, for a certain project development, the terminal device may determine a plurality of tasks to be performed and a plurality of task execution groups in advance. After the terminal device obtains the plurality of task execution subgroups and the plurality of tasks to be executed, task feature sequences corresponding to the plurality of tasks to be executed and task subgroup sequences of the plurality of task execution subgroups can be obtained, and further task matching weights between the task subgroup sequences and the task feature sequences can be determined through a preset neural network. Furthermore, the terminal device can determine task dependency relationships among a plurality of tasks to be executed, determine resources required by the task execution group to execute each task to be executed, and construct a resource demand constraint matrix. And acquiring an optimal task allocation scheme through a genetic algorithm according to the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining the task execution time of each task to be executed.
Compared with the task allocation mode in the prior art, the task allocation optimal scheme can be solved through a preset neural network, the task matching weight between the task group sequence and the task feature sequence, the task dependency relationship among a plurality of tasks to be executed, the resource demand constraint matrix when each task execution group executes the task to be executed and the task execution time corresponding to the task to be executed. Therefore, the invention considers the constraint relation between the task execution group and the task to be executed, the constraint among the tasks to be executed, the constraint of resources required by the task to be executed and the constraint of the task execution time, and determines the optimal task allocation scheme through the genetic algorithm under a plurality of constraint conditions, so that the optimal task allocation scheme can accurately adapt to the task execution capacity and the resource requirement of each group, the task accurate allocation is realized, and the subsequent task processing efficiency is further improved.
Further, based on the first embodiment of the task allocation of the present invention, a second embodiment of the task allocation of the present invention is presented.
In this embodiment, in the step S20, before "obtaining the task feature sequences of the plurality of tasks to be executed and the task group sequences of the plurality of task execution groups, and determining, by a preset neural network, the task matching weights between the task group sequences and the task feature sequences, the method may further include:
step S50, aiming at each task to be executed in the current time window, acquiring task weight of the task to be executed;
step S60, constructing the preset neural network according to the sequence sizes of the task feature sequence and the task group sequence;
in this embodiment, according to the above description, after the terminal device collects the historical performance data of m tasks in a data collection window, a task execution time matrix E may be constructed nm Task completion quality matrix Q nm Group resource utilization vector U n Overdue days vector D m Then, the task weight W of each task to be executed (hereinafter referred to as task) can be calculated ij ＝T ij *U i *P j Wherein, the method comprises the steps of, wherein,scoring task execution efficiency, U i ＝U i Is a resource utilization factor->Is the overdue penalty factor (α is used to control the effect of overdue days on penalty factor).
Further, a vector with an input layer of 1 x (x+y) is constructed, and a neural network with an output layer of a single node is initialized. Wherein x is the sequence size of the task feature sequence, and y is the sequence size of the task group sequence.
On this basis, in the step S20, the "determining, by the preset neural network, the task matching weight between the task group sequence and the task feature sequence" may include:
step S201, mapping the task feature sequence and the task group sequence into boolean vectors with equal lengths, and using the boolean vectors as input of the preset neural network to obtain a task matching weight between the task group sequence and the task feature sequence output by the preset neural network, where a significance probability between the task matching weight and the task weight of the task to be executed is smaller than a preset difference threshold.
Step S202, if the saliency probability between the task matching weight and the task weight is greater than the preset difference threshold, training the preset neural network by a back propagation mode until the saliency probability between the task matching weight and the task weight is less than the preset difference threshold.
In this embodiment, the terminal device acquires the task weight and the configuration of the task to be executedAfter the neural network corresponding to each group is built, the task feature sequence can be obtainedAnd task team sequence->And respectively mapping the task matching weights into equal-length Boolean vectors, and taking the Boolean vectors as the input of the neural network to obtain task matching weights between task group sequences and task feature sequences output by the neural network, wherein the probability of the significant rate between the task matching weights and the task weights is smaller than a preset difference threshold.
Specifically, for example, the terminal device may perform T-test on the task weight in the last data acquisition window and the output value list corresponding to the neural network (the output value list includes the task matching weights of the group for each task), and when the difference is not significant (for example, the probability of the significance ratio between the task matching weight and the task weight is less than a preset difference threshold), update no neural network parameter and update the data acquisition window interval to 2τ.
When the difference is considered to be significant (for example, the difference between the task matching weight and the task weight is greater than or equal to a preset difference threshold), parameter optimization is performed on each neural network in a back propagation mode, so that self-learning is performed on the actual environment until the probability of the significant rate is smaller than the preset difference threshold.
For example, group i has a task matching weight of W for task j, which has a task weight of W ij If W is ij The probability of the saliency ratio between the neural network parameter and W is smaller than a preset difference threshold (for example, 0.5 can be taken), and the neural network parameter does not need to be updated.
In this embodiment, the value of the preset difference threshold is not specifically limited.
Further, as shown in fig. 4, the fitness function may be constructed for the genetic algorithm based on the task weights output from the neural network.
Thus, in this embodiment, tasks and task groups are abstracted into non-subdividable and independent features, and based thereon, neural networks describing the relationships between features and tasks are constructed. When a new task or a task group is generated and changed, the actual weight value can be quickly fitted through the neural network only by carrying out characteristic decomposition. The adaptive degree weight parameters of self-learning are introduced through data acquisition, data preprocessing, task characterization and neural network construction, and the starting of neural network training is controlled based on significance analysis and a time window, so that the adaptive degree parameters in a genetic algorithm can be effectively updated according to the specific performances of the genetic algorithm when the system is in operation, so that the dynamic change of tasks and teams is adapted, the rationality of task allocation is ensured, and the task processing efficiency is improved.
Further, based on the first and second embodiments of the task allocation of the present invention, a third embodiment of the task allocation of the present invention is proposed.
In this embodiment, in the above step S30, "determining the task dependency relationship between the plurality of tasks to be executed" may include:
step S301, obtaining a dependency relationship between the plurality of tasks to be executed;
step S302, a task dependency graph is constructed according to the dependency relationship, wherein each node of the task dependency graph represents each task to be executed, and the directed edge of the task dependency graph represents the dependency relationship.
In this embodiment, the dependency relationship among the tasks may be understood as task T a To split a certain system, task T b To optimize a certain system, the optimization can be after splitting, so T b The start of (1) depends on T a Is completed.
On the basis, after acquiring a plurality of tasks, the terminal equipment can acquire the dependency relationship among the plurality of tasks to be executed, and further, can construct a task dependency graph according to the dependency relationship, wherein each node of the task dependency graph comprises each task to be executed, and the directed edges of the task dependency graph represent the dependency relationship.
Specifically, for example, the terminal device may establish a task dependency graph G, whichIn the above, the nodes represent tasks, and the directed edges represent the dependency relationships between the tasks. For example, if it is task T a And T b If the task has a dependency relationship, a slave node T exists in the task dependency graph G a Pointing node T b Is a directional edge of (a).
Further, in the above step S30, "constructing a resource requirement constraint matrix when the task execution team executes the task to be executed" may include:
step S303, obtaining the resource demand of the task execution group to execute the task to be executed;
step S304, constructing a resource demand constraint matrix according to the resource demand, wherein the resource demand constraint matrix comprises the resource demand of each task to be executed by each task execution group.
In this embodiment, the terminal device may construct the resource requirement constraint matrix R NM Where N represents the number of subgroups and M represents the number of tasks. R is R ij Representing the resource demand of the ith team in performing the jth task. In task allocation, R needs to be satisfied ij ≤A i ，A i Indicating the amount of available resources for the ith subgroup.
For example, the i-th group is estimated to require 2.5 people/day when performing the j-task, but in practice only 2 people/day of the i-th group Ai can be provided for the corresponding time, the task cannot be assigned to the group.
Therefore, in the present embodiment, the accuracy of evaluation of the merits of the scheme is enhanced by performing fitness evaluation by the above-described various constraints, such as the constraints between the group and the task, the constraints between the group and the task.
Further, a fourth embodiment of the task allocation of the present invention is proposed based on the first, second and third embodiments of the task allocation of the present invention.
In this embodiment, before "obtaining the task allocation and scheduling optimal scheme by genetic algorithm based on the task dependency relationship, the resource requirement constraint matrix, and the task matching weight and in combination with the task execution time corresponding to each task execution group" in step S40, the method may further include:
step S70, obtaining dependency relation weights according to the task dependency graph;
step S80, acquiring inter-group resource load balancing weights corresponding to the task execution groups according to the resource demand constraint matrix;
step S90, superposing the task matching weight, the dependency relation weight and the inter-group resource load balancing weight to obtain an adaptability function;
step S100, constructing an initialized population, wherein each individual in the initialized population represents a task allocation scheme;
In this embodiment, as shown in fig. 4, the terminal device needs to construct an initialization population in the genetic algorithm in advance.
Specifically, for example, if there are N subgroups, M tasks. A population of size N is initialized, comprising N randomly generated individuals, each individual representing a task allocation scheme, represented by chromosomes of length M, each individual representing which subgroup the task is allocated to.
Furthermore, the terminal device needs to construct the fitness function in advance before searching for the optimal allocation scheme through the genetic algorithm.
Specifically, for example, the terminal device may define a comprehensive fitness function to evaluate the merits of individuals in the genetic algorithm, where the fitness function is derived by superposition of three parts:
(1) Considering the matching degree between the group and the task, namely the task matching weight output by the neural network;
(2) Considering the dependency relationship between tasks, that is, the dependency relationship weight between at least two tasks calculated by the task dependency graph G, the existing directed graph weight solving manner can be referred to, and details are not repeated here;
(3) Considering inter-group resource load balancing, i.e. constraint matrix R by resource demand NM Calculating to obtain inter-group resource load balancing weights corresponding to the groups, wherein the inter-group resource load balancing weights can be processed Solving the task assigned between the teams will reach a relative balance, avoiding the team from working overload.
The calculation of the inter-group resource load balancing weight specifically comprises the following steps:
constraint matrix R for resource requirements NM The value of (2) is equal to A i To obtain the Boolean matrix W NM When R is ij ≤A i At the time W ij =1, otherwise W ij ＝0；
Calculate matrix W NM Sum of each row of (a)
Finally, the inter-group resource load balancing weight isWherein (1)>
Furthermore, the terminal device may obtain the task matching weight, the dependency relationship weight, and the inter-group resource load balancing weight, and then perform weight superposition on the task matching weight, the dependency relationship weight, and the inter-group resource load balancing weight, to obtain the fitness function.
Further, in the step S40, the "obtaining, by genetic algorithm, the optimal task allocation and scheduling scheme based on the task dependency relationship, the resource requirement constraint matrix, and the task matching weight and in combination with the task execution time corresponding to each task execution group" may include:
step S401, determining a task set to be executed corresponding to each task execution group;
step S402, acquiring task time limits of each task to be executed in the task set to be executed, wherein the task time limits comprise task execution time and task execution deadline;
Step S403, calculating an urgent value of each task to be executed according to the task execution time and the task execution deadline, and combining the current time;
and step S404, optimizing the task allocation scheme through a genetic algorithm according to the fitness function, and sequencing a plurality of tasks to be executed according to the urgent value in the optimizing process until the task allocation optimal scheme is obtained.
In this embodiment, after the initialization function and the fitness function are constructed, as shown in fig. 4, the terminal device performs the sorting in the time dimension by adopting an algorithm of scheduling according to the urgency (CR) value for the task set inside the group corresponding to each chromosome. The algorithm may consider the urgency of the task to satisfy the time constraint of the task:
(1) Determining a task set to be executed corresponding to each task execution group; the task processing time of the task i in the task set to be executed is p i The task execution deadline is d i And calculating the elapsed time t in combination with the current time, and further calculating the CR value of each task:
(2) Sequencing the tasks in the task set to be executed according to the CR value: sequencing all tasks according to the CR value obtained by calculation in order from small to large;
(3) And (3) adaptability evaluation: according to the fitness function, performing fitness evaluation on each individual in the population, and measuring the performance of the scheduling scheme corresponding to the individual on the problem;
(4) Selection, crossover and mutation: adopting a roulette selection strategy, generating father generation preferentially according to the performances of each individual during fitness evaluation, and generating new individuals through multi-point crossing and single-gene variation so as to maintain the diversity of population and increase the search space of potential solutions;
(5) Population updating: updating the population to a combination of newly generated individuals and a portion of the parent individuals;
(6) Iterative optimization: repeating the steps (1) to (5) to improve the overall fitness of the population.
(7) Outputting the optimal individual and decoding: when the termination condition is met (for example, the iteration times reach a preset value), taking the individual with the highest fitness in the final population as an optimal solution, and acquiring a task allocation scheme corresponding to the individual with the highest fitness, namely, the task allocation optimal scheme.
Therefore, in this embodiment, a dual-layer scheduling policy is adopted, that is, tasks are firstly allocated to each subgroup through crossover and mutation operations in a genetic algorithm, resource allocation is optimized, constraint relationships between the tasks and the subgroups are considered, and then, for the tasks in each subgroup, the intra-group scheduling optimization algorithm is used for performing time-dimensional ordering to meet the time constraint, so that the problems of large search space, long convergence time and the like of the genetic algorithm are avoided, the rationality of task allocation is ensured, the subgroups can finish the tasks according to the task, and further, the task processing efficiency is improved.
In general, the invention relates to a multi-task constraint scheduling method and system based on a self-learning genetic algorithm, which are used for optimizing and configuring development resources of different groups in a large-scale software development team. The method fully considers the efficiency difference of different groups when developing different tasks, and adapts to the dynamic changes of the tasks and teams by the parameters of each group for developing the tasks in a self-learning mechanism real-time algorithm. Meanwhile, the multi-task constraint is comprehensively considered, and not only comprises constraint relation between the group and the task, but also covers constraint between the task and the group. In addition, in solving the problems of large search space and long convergence time in the conventional technical scheme, the invention adopts double-layer rank Cheng Celve, performs task allocation through a genetic algorithm, and then performs ranking in time dimension through an intra-work-group scheduling optimization algorithm. Finally, the convergence direction is controlled through the comprehensive fitness function, so that a stable and efficient solution is found.
In addition, an embodiment of the present invention further provides a task allocation device, referring to fig. 5, where the task allocation device includes:
a first determining module 10, configured to determine a plurality of tasks to be performed and a plurality of task execution groups;
The second determining module 20 is configured to obtain task feature sequences of the plurality of tasks to be executed and task subgroup sequences of the plurality of task execution subgroups, and determine task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network;
a third determining module 30, configured to determine task dependencies among the plurality of tasks to be executed, and construct a resource requirement constraint matrix when the task execution team executes the tasks to be executed;
the obtaining module 40 is configured to obtain a task allocation optimal solution based on the task dependency relationship, the resource requirement constraint matrix, and the task matching weight, and in combination with a task time limit corresponding to each task to be executed.
The expansion content of the specific implementation mode of the task allocation device is basically the same as that of each embodiment of the task allocation method, and is not described in detail herein.
In addition, the embodiment of the invention also provides a computer readable storage medium, wherein the storage medium stores a task allocation program, and the task allocation program realizes the steps of a task allocation method when being executed by a processor.
Embodiments of the task allocation apparatus and the computer readable storage medium of the present invention may refer to embodiments of the task allocation method of the present invention, and will not be described herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal device (which may be a wearable device, a locator, a smart phone, a tablet computer, etc.) to perform the method according to the embodiments of the present invention.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (10)
1. A task allocation method, characterized in that the task allocation method comprises:
determining a plurality of tasks to be executed and a plurality of task execution groups;
acquiring task feature sequences of the tasks to be executed and task subgroup sequences of the task execution subgroups, and determining task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network;
determining task dependency relations among the plurality of tasks to be executed, and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed;
and acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed.
2. The task allocation method according to claim 1, further comprising, before the step of acquiring task feature sequences of the plurality of tasks to be executed and task group sequences of the plurality of task execution groups, determining task matching weights between the task group sequences and the task feature sequences through a preset neural network:
Acquiring task weights of the tasks to be executed aiming at each task to be executed in a current time window;
constructing the preset neural network according to the sequence sizes of the task feature sequence and the task group sequence;
the step of determining the task matching weight between the task group sequence and the task feature sequence through a preset neural network comprises the following steps:
and mapping the task feature sequence and the task group sequence into equal-length Boolean vectors respectively, and taking the Boolean vectors as the input of the preset neural network to obtain task matching weights between the task group sequence and the task feature sequence output by the preset neural network, wherein the significance probability between the task matching weights and the task weights of the tasks to be executed is smaller than a preset difference threshold.
3. The task allocation method according to claim 1, further comprising, after the step of mapping the task feature sequences into boolean vectors of equal length and taking the boolean vectors as inputs of the preset neural network, obtaining task matching weights between the task execution subgroups output by the preset neural network and the task feature sequences of the tasks to be executed:
If the significance probability between the task matching weight and the task weight is larger than the preset difference threshold, training the preset neural network in a back propagation mode until the significance probability between the task matching weight and the task weight is smaller than the preset difference threshold.
4. The task allocation method according to claim 1, wherein the step of determining task dependencies among the plurality of tasks to be executed comprises:
acquiring the dependency relationship among the plurality of tasks to be executed;
and constructing a task dependency graph according to the dependency relationship, wherein each node of the task dependency graph represents each task to be executed, and the directed edge of the task dependency graph represents the dependency relationship.
5. The task allocation method according to claim 4, wherein the step of constructing a resource requirement constraint matrix when the task execution team executes the task to be executed, comprises:
acquiring the resource demand of the task execution group to execute the task to be executed;
and constructing a resource demand constraint matrix according to the resource demand, wherein the resource demand constraint matrix comprises the resource demand of each task to be executed for executing each task of each task execution group.
6. The method for allocating tasks according to claim 5, further comprising, before the step of obtaining an optimal task allocation solution based on the task dependency relationship, the resource requirement constraint matrix, and the task matching weights and in combination with task time limits corresponding to each task to be executed:
constructing an initialized population, wherein each individual in the initialized population characterizes a task allocation scheme;
acquiring dependency relation weight according to the task dependency graph;
acquiring inter-group resource load balancing weights corresponding to the task execution groups according to the resource demand constraint matrix;
and carrying out weight superposition on the task matching weight, the dependency relationship weight and the inter-group resource load balancing weight to obtain an adaptability function, so as to obtain a task allocation and scheduling optimal scheme by combining task execution group corresponding task time limits according to the adaptability function and through a genetic algorithm.
7. The task allocation method according to claim 6, wherein the step of obtaining the optimal task allocation scheme based on the task dependency relationship, the resource requirement constraint matrix, and the task matching weights and in combination with the task time limit corresponding to each task to be executed includes:
Determining a task set to be executed corresponding to each task execution group;
acquiring task time limits of all the tasks to be executed in the task set to be executed, wherein the task time limits comprise task execution time and task execution deadline;
calculating the urgent value of each task to be executed according to the task execution time and the task execution deadline and by combining the current time;
and optimizing the task allocation scheme through a genetic algorithm according to the fitness function, and sequencing a plurality of tasks to be executed according to the urgent value in the optimizing process until the task allocation optimal scheme is obtained.
8. A task allocation device, characterized in that the task allocation device comprises:
the first determining module is used for determining a plurality of tasks to be executed and a plurality of task execution groups;
the second determining module is used for acquiring task feature sequences of the plurality of tasks to be executed and task subgroup sequences of the plurality of task execution subgroups, and determining task matching weights between the task subgroup sequences and the task feature sequences through a preset neural network;
the third determining module is used for determining task dependency relations among the plurality of tasks to be executed and constructing a resource demand constraint matrix when the task execution group executes the tasks to be executed;
The acquisition module is used for acquiring a task allocation optimal scheme based on the task dependency relationship, the resource demand constraint matrix and the task matching weight and combining task time limits corresponding to each task to be executed.
9. A terminal device comprising a memory, a processor and a base task allocation program stored on the memory and executable on the processor, the task allocation program when executed by the processor implementing the steps of the task allocation method according to any one of claims 1 to 6.
10. A computer-readable storage medium, on which a task allocation program is stored, which when executed by a processor, implements the steps of the task allocation method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311085576.8A CN117436627A (en) | 2023-08-25 | 2023-08-25 | Task allocation method, device, terminal equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311085576.8A CN117436627A (en) | 2023-08-25 | 2023-08-25 | Task allocation method, device, terminal equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117436627A true CN117436627A (en) | 2024-01-23 |
Family
ID=89548737
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311085576.8A Pending CN117436627A (en) | 2023-08-25 | 2023-08-25 | Task allocation method, device, terminal equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117436627A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118333277A (en) * | 2024-04-29 | 2024-07-12 | 北京中核华辉科技发展有限公司 | Scheduling system and method based on nuclear engineering work plan |
-
2023
- 2023-08-25 CN CN202311085576.8A patent/CN117436627A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118333277A (en) * | 2024-04-29 | 2024-07-12 | 北京中核华辉科技发展有限公司 | Scheduling system and method based on nuclear engineering work plan |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Abed-Alguni et al. | Distributed Grey Wolf Optimizer for scheduling of workflow applications in cloud environments | |
| Han et al. | Tailored learning-based scheduling for kubernetes-oriented edge-cloud system | |
| Kaur et al. | Deep‐Q learning‐based heterogeneous earliest finish time scheduling algorithm for scientific workflows in cloud | |
| Jayanetti et al. | Deep reinforcement learning for energy and time optimized scheduling of precedence-constrained tasks in edge–cloud computing environments | |
| EP3938963A1 (en) | Scheduling computation graphs using neural networks | |
| Xiao et al. | A cooperative coevolution hyper-heuristic framework for workflow scheduling problem | |
| US20150170052A1 (en) | Method of reducing resource fluctuations in resource leveling | |
| CN109491761A (en) | Cloud computing multiple target method for scheduling task based on EDA-GA hybrid algorithm | |
| CN112784362A (en) | Hybrid optimization method and system for unmanned aerial vehicle-assisted edge calculation | |
| CN113821318B (en) | Internet of things cross-domain subtask combination collaborative computing method and system | |
| CN114580678A (en) | Product maintenance resource scheduling method and system | |
| CN117436627A (en) | Task allocation method, device, terminal equipment and medium | |
| Subramoney et al. | A comparative evaluation of population-based optimization algorithms for workflow scheduling in cloud-fog environments | |
| CN111343259B (en) | Binary code-based cloud task scheduling method, server and storage medium | |
| Mohammadzadeh et al. | Energy-aware workflow scheduling in fog computing using a hybrid chaotic algorithm | |
| CN117271101A (en) | Operator fusion method and device, electronic equipment and storage medium | |
| CN116033026A (en) | Resource scheduling method | |
| CN116257363A (en) | Resource scheduling method, device, equipment and storage medium | |
| Entezari-Maleki et al. | A genetic algorithm to increase the throughput of the computational grids | |
| CN115794323A (en) | Task scheduling method, device, server and storage medium | |
| CN113504998A (en) | Method, device and equipment for determining task scheduling scheme | |
| CN111258743A (en) | Cloud task scheduling method, device, equipment and storage medium based on discrete coding | |
| Lin et al. | Runtime estimation and scheduling on parallel processing supercomputers via instance-based learning and swarm intelligence | |
| CN114980216A (en) | Dependent task unloading system and method based on mobile edge calculation | |
| Ebadifard et al. | A multi-objective approach with waspas decision-making for workflow scheduling in cloud environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |