Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Parallel cloud workflow scheduling method based on reinforcement learning strategy

A technology of reinforcement learning and workflow, applied in the field of cloud computing, can solve the problems that the neural network cannot handle the input information of dimension changes, cannot learn the knowledge of the task to be selected, and it is difficult to perceive dynamic changes, etc., so as to optimize the action selection strategy and increase Variety, the effect of improving optimization performance

Active Publication Date: 2021-01-15
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, these workflow scheduling methods based on reinforcement learning strategies have the following shortcomings: the traditional neural network used by the agent cannot handle the input information of dimension changes, and it is difficult to perceive the dynamic changes in the number of tasks to be selected in different stages of workflow scheduling, that is, Unable to learn the knowledge about the tasks to be selected, which directly affects scheduling decisions, such as task selection and task-to-resource mapping

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Parallel cloud workflow scheduling method based on reinforcement learning strategy
  • Parallel cloud workflow scheduling method based on reinforcement learning strategy
  • Parallel cloud workflow scheduling method based on reinforcement learning strategy

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0048] The present invention will be described in detail below with reference to the accompanying drawings and examples.

[0049] In the prior art, the standard reinforcement learning algorithm DQN includes two parts, the Agent and the environment, and uses a neural network to approximate the action state value function. Among them, the interaction process between the Agent and the environment is as follows: In time step t, firstly, the Agent receives the state information of the environment (s t ), choose the action to be taken on the environment (a t ); Secondly, the action a t Act on the environment and get the environment's reward for the action (r t+1 ) and the next state after the environment update (s t+1 ); In the next time step t+1, first, judge whether the round is terminated. If the round is not over, the Agent obtains new environmental state information and performs new interactions with the environment. The specific framework is as follows figure 2 shown. D...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a parallel cloud workflow scheduling method based on a reinforcement learning strategy, and the method comprises the steps: introducing a pointer network in a task selection process, taking softmax probability distribution as a pointer for processing variable-length input so that a workflow scheduling model is capable of sensing the dynamic change of a to-be-selected task at different workflow scheduling stages, thus improving the task selection efficiency. More task execution sequence knowledge is learned, and the optimization performance of a scheduling solution is improved.

Description

technical field [0001] The invention belongs to the technical field of cloud computing, and in particular relates to a parallel cloud workflow scheduling method based on a reinforcement learning strategy. Background technique [0002] As a new type of computing service provision model, cloud computing has the characteristics of flexible resource allocation and pay-per-use, and can provide network users with flexible, efficient, and scalable computing and storage resource services without space and time constraints. Users can obtain computing, storage and other services through the network without purchasing hardware resources such as servers. With the rapid development of cloud computing and the increasing size and complexity of scientific application data, more and more large-scale scientific applications are deployed or are being migrated to cloud data centers for execution. The continuous expansion of cloud applications and the increasing number of cloud user requests ha...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F9/48G06N3/04G06N20/00
CPCG06F9/4881G06N20/00G06N3/045Y02D10/00
Inventor 李慧芳黄姜杭王彬阳王一竹王丹敬邹伟东柴森春夏元清
Owner BEIJING INSTITUTE OF TECHNOLOGYGY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com