The Three-Motor Synchronous Control System Based On ADRC With RBF Neural Network

With the rapid development of science and technology,industrial automation has gradually penetrated into all walks of life.The multi-motor control with high efficiency,performance and economy has become a hot discussion.In this paper,a series of research on the controller parameters optimization and the decoupling problem of three-motor synchronous control is carried out.Based on analyzing the model of three-motor synchronous control system,the decoupling problem between tension and speed was put forward.Due to the shortage of the traditional PID control scheme,a new control strategy of the three-motor synchronous control system was designed based on the first-order active disturbance rejection controller(ADRC)according to the independence of specific model and the external and internal disturbance as the whole disturbance to be observed and compensated,and then compared with the PID control strategy by simulation experiments.The results show that the ADRC is superior to PID in adjusting speed,overshoot and anti-interference.Secondly,to solve the complex algorithm and the difficulty of adjusting parameters of ADRC,an adaptive ADRC was designed based on radial basis function neural network(RBFNN).The controller uses the RBFNN can approximate any nonlinear function with arbitrary precision and simple structure to realize the online tracking of the controlled object,then the on-line adjusting information of ADRC parameters obtained to realize self-adjusting function of ADRC parameters.The control system used programmable logic controller(PLC)S7-300 as the controller,The PROFINET communication is used to connect the host computer to complete the hardware configuration,program download,debugging and other functions.At the same time,a remote monitoring system is designed by the Windows Control Center(WinCC)configuration software to realize real-time monitoring and observe control process change.The PROFIBUS-DP communication is used to connect converters to complete the contruction of master-slave communication network.The programs are written by the method of structured programming which using Portal V13 software in the host computer,such as ADRC,parameter acquisition,communication,RBFNN and so on.The experiments are performed on the above experimental platform,including the step response experiment,decoupling experiment and anti-interference experiment of the adaptive ADRC based on RBFNN.The results show that RBFNN has the advantages in nonlinear mapping,show the adaptive ADRC can not only achieve the parameters’ self-adjusting function,but also can decrease the overshoot,improve the dynamic performance and steady state precision of the system.The experimental results show that the method is practical.