| The fourth industrial revolution requires more flexibility,portability,configurability,and miniaturization of industrial control systems.However,the software and hardware of the industrial controller are integrated,and its software definition capability and migration scheduling capability are insufficient.In this paper,the new type of industrial controller is studied for practical problems.A software-defined industrial controller design scheme and an industrial task and network scheduling method based on the controller are proposed.Then the feasibility and effectiveness of key technologies are verified.The main contributions of this paper include the following aspects:A software-defined industrial controller design scheme is proposed.This scheme uses hardware virtualization technology,container technology and software-defined technology to separate the software/hardware of the industrial control system.Then,this scheme realizes location-independent data access through I/O interface virtualization and access interface standardization and improves the software-defined capability and migration scheduling capability of industrial controllers.The proposed software-defined industrial controller can construct field control devices and data acquisition and monitoring devices with different roles by loading different control application software.It realizes the design goal of determining the role and control function of the devices by software,that is,the software-defined industrial controller.Software-defined industrial controllers can effectively reduce the type and quantity of hardware devices,making the deployment and maintenance of industrial control systems more convenient.Moreover,it can protect the existing investment,and the cost of system upgrading is also lower.Finally,a series of experiments are designed to verify the technical feasibility of the proposed software-defined industrial controller.Based on the software-defined industrial controller,the task and network scheduling of distributed industrial applications are studied.An industrial task scheduling method based on multi-objective optimization is proposed.This method utilizes multiple software-defined industrial controllers to build a dispersed control system,breaking the traditional "end-to-end" computing paradigm.It adopts the computing paradigm of "hop-by-hop forwarding and computing".Industrial data is processed hop-by-hop during the hop-by-hop forwarding of the control network.Aiming at the scheduling problem of industrial tasks in a dispersed computing environment,this method models industrial task deployment and network flow scheduling as a multi-objective optimization problem.Then,this optimization problem is solved using the genetic algorithm.And experiments are designed to verify the effectiveness of the proposed scheduling method.The software-defined industrial control prototype device-u Tesla designed and implemented by our research group is described in detail.The device uses Xvisor as hardware virtualization.A distributed control application based on IEC 61499 is built using the Eclipse4 diac open-source project.The application's functional blocks and runtime environment Forte are packaged into Docker containers and deployed to u Tesla devices for execution.The research results of this paper have good theoretical value and practical significance for the development of industrial control system and provide important support for the construction of industrial control system in the era of industry 4.0. |
PDF Full Text Request