Research On Synchronous Control And Scheduling Cooperative Design In Networked Distributed Motion Control System

In the recent years,with the increasing degree of networking,intelligence and informationization of the advanced industrial manufacturing systems,the distributed networked multiaxial motion control systems(DNMAMCs)have become the advanced mark of the manufacturing systems.As the network scale increases,however,the design of such systems becomes more and more complicated.These systems not only need to consider network induction issues such as time delay,jitter,packet loss,and timing disorder,but also need to carefully consider the coordination of communication task scheduling and multi-axis motion control between network nodes.Thus,resource sharing and coordinated operation among distributed networked nodes are realized.In this paper,the studies base on the overall design framework which established according to a typical three-layer structure of the industrial manufacturing system which,specifically,at the system operation layer,it provides DNMAMCs system application management and control,including system task planning(such as service selection,service composition and coordination),system operation and maintenance(such as system performance monitoring and fault detection,etc.);at the network transport layer,it control the time synchronization between different communication tasks along the data packet transmission path(such as status feedback data collection,control command issuance,etc.);at the device level,sensors and controllers are used to achieve a certain specific real-time sensing and dynamic control tasks.This paper mainly focus on two core issues for designing a DNMAMCs system: one is the issue of the distributed network communication cooperative scheduling and the other is the issue of multi-axis motion synchronous control.The related research content is summarized as follows:(1)Aiming at the problems of long transmission path delay,delay jitter,and channel asymmetry in the multi-hop communication structure of DNMAMCs systems,a network communication task coordination method based on distributed clock and time synchronization technology is proposed by this paper.Firstly,this method uses the transparent clock method defined in the IEEE1588 PTP protocol to establish a distributed clock,and thus a unified time metric can be obtained by the DNMAMCs system.Secondly,based on the time division multiple access(TDMA)mechanism,a scheduling schedule can be generated for scheduling communication tasks of the DNMAMCs system by assigning the Guaranteed Time Slots(GTS)to distributed network nodes,which are used to generate the communication task control timing of each network node.As a result,this method can unify the hierarchical control network of the DNMAMCs system into time-triggered real-time Ethernet,and the DNMAMCs system is decomposed into decompose a set of delay-related parallel distributed periodic subsystems.Finally,a cascade controller based on a Kalman filter and a PI servo controller is designed for improving the performance of time synchronization.(2)Aiming at the problem of asynchronous coordination between status feedback and control command issuance in each motion control loop under the framework of GTS communication task scheduling schedule,a double-layer architecture based on virtual coding and a parallel distributed network control scheme are proposed.First,according to the motion task planning of the DNMAMCs application system,the overall performance index is decomposed into a series of local control loop expected output reference values which are used as the setpoints of each local controller.These decomposed setpoints can be regarded as the reference trajectory of local control loops,thereby,the motion control problem at the device level can be transformed into the output tracking control problem.Secondly,at the operation and device layer,a TS-type fuzzy neural network predictive controller and an adaptive PID controller based on RBF fuzzy neural network is developed by DNMAMCs system.This type of predictive controller is used as the successive state estimator of the DNMAMCs system,which is used to update the setpoint and corresponding control parameters of the PID controller output at the equipment layer.The predictive controller successively estimates the operating conditions of each motion axis,updates the setting value and control parameters of the PID controller,then the PID controller implements the control action to the corresponding motion axis system.In this paper,a synchronous motion control and communication scheduling collaborative design method which integrates clock information into control process is proposed.By analyzing and calculating the response time of communication tasks,network quality of service(Qo S)parameters such as data transmission delay and jitter can be used as the design parameters for network controller,enabling designers to better analyze and understand the dynamics of DNMAMCs system,providing a design framework and implementation method for establishing a DNMAMCs system integration model that integrates network communication,sub-calculation,and collaborative control.Under the framework of time synchronization communication scheduling,according to the cross-level task scheduling table,a parallel distributed double-layer structure based on virtual coding is adopted to construct a predictive controller and a PID motion controller,coordinate the operation condition monitoring in management layer and motion axis control in execution layer.Thus,through high-precision followup control and synchronization effects of each motion axis,the coordinated consistency of multi-axis motion control can be obtained.Finally,three experiments of time synchronization accuracy under multi-hop communication structure single axis tracking control based on virtual axis,triple axis trajectory synchronization control physical simulation were carried out,inspecting the sampling period,control cycle stability,performance and multi-axis motion axis follow movement synchronization ability respectively.Results show the effectiveness of the proposed method.