Research On Networked Control System Platform Development And Data-driven Control Algorithms

In the process of developing the control system,the conventional method of realizing the control algorithm by software is manual programming.As programmers need to debug and verify several times after programming,the development cost-control,the efficiency,the reliability and the quality of application needs improvement.Based on the real-time control requirements and the technical characteristics of the hardware-in-the-loop simulation platform,this paper uses Simulink Real-Time and Lab VIEW to build the real-time control platform.A series of custom S-functions are designed to realize the real-time control function.As an example,the position control of the servo motor and the network communication on this platform is implemented,two kinds of data-driven predictive control algorithms for the problem of delay and packet loss in networked control has been made out.Firstly,the design of networked control system platform and data driven algorithm is proposed,including platform environment setting,hardware selection,motor control,network communication,delay compensation algorithm and simulation.Secondly,it discusses the construction of real-time control platform based on Simulink Real-Time,and the key technologies and methods to control the position of the motor and realize the communication with LabVIEW.To maximize the use of the motor Modbus communication protocol communication interval,the design of a series of custom s-function to achieve real-time motor position control,compile the file from the LabVIEW function by self-encapsulation DLL files based on Simulink,and achieve the function that two targets communicate with each other through Ethernet and position control of the servo motor under the PID algorithm.Thirdly,a data-driven predictive control method is proposed for the network delay and packet loss and out-of-order in networked control.It can effectively alleviate the uncertainties and disturbances caused by the network environment.Then the simulation and platform experiments show that the data-driven predictive control algorithm is reliable and effective in compensating network induced delay and packet loss and outlier problem.Finally,the data-driven predictive control method for the network with time delay and packet loss is improved by using the system input selected in the sliding window for several consecutive times before it can be distributed.It has a stronger ability to track time-varying parameters,system's complex dynamic behavior and has a better dynamic performance.And then through the simulation based on the network control platform in this article I verify effectiveness of the algorithm.