Research And Implementation Of Cloud Control System

With the development of control theory gone through classical control and modern con-trol,even to today's intelligent control,the plants grow Increasingly with more information involved and control algorithms becoming more and more complex.Due to the limitation of software and hardware,the conventional control system has no enough ability to deal with complex systems and run complex algorithms.Over the past decade,Cloud Computing and Big Data developed in Internet field and are getting mature,which makes it's feasible to process complex missions and mass data.Thus,Internet of Things(IoT)is getting hotter and walking toward ground implementation.The application of these new technologies to control theory brings it back to life and puts foward the concept of Cloud Control System(CCS)In order to exploit problems of CCS and speed up its engineering implementation,the paper works on research and implementation of CCS.The main contents are as follows:?The concept of CCS.It describes the concept of CCS from the view of control and computing,including the idea of "Control as a Service",Initial Phase and Cloud Con-trol Phase.The ralations of cloud control,cloud computing and edge computing are discussed with edge control enriching the connotation of CCS?Design of Intial Phase in CCS.The key of engineering technology lies on solving practical problems.The paper sets up a series of experiments to test detail parameters of network communication in Initial Phase and provides a simple solution.It applies in-cremental PID(proportional-integral-derivative)to ball-beam system to realize a ball-beam CCS and makes exploration on effectiveness and limitations of the solution?Subspace predictive control in CCS.When systems and network get more complex,PID shows no performance advantage.Subspace predictive control is used.On the one hand,it only asks for output and input informations of system,on the other,predictive control could solve large network delay to a great extent.The paper makes detail description of subspace theory and subspace predictive control.It applies subspace predictive control to ball-beam CCS and shows that it's effective in dealing with long network-induced time-delay?Cloud control service platform architecture.For the sake of real "Control as a Ser-vice",it ends with an 4-layers achitecture of cloud control,namely,management layer,service layer,network layer and edge layer,with modules of monitoring and security management.This architecture may work as a reference in implementation of cloud control platform.Also,a rudiment of platform is proposed,which may accelerates the research and development of cloud control platformThe results of this paper may make a relatively strong reference significance to the further research on CCS.It solves some engineering problems and may show good demon-stration of and guidance on how to implement cloud control.