Sampling time scheduling for a CAN-based dynamical system: A simulation practice

This paper addresses the method to obtain maximum allowable loop delay (MALD) for calculation of data sampling time with offset scheduling (DSTOS) for networked dynamic system with multiple control loops via controller area network (CAN). The DSTOS algorithm calculation is based on MALD that is obtained by simulation means in Matlab/SimuUnk. The step-by-step of MALD selection based simulation is presented. A networked control system of direct-drive 2-DOF SCARA Robot is developed based on CAN where PD controller is adapted to form the closed loop system. The simulation is performed by using Matiab/Simulink-based TrueTime Toolbox and the implementation of DSTOS for three SCARA robot via CAN is analysed in event of the data latency of control data from sensor to actuator and the performance index, integral square error (IAE) of the link angle. The results and analysis show that the proper selection of sampling time for each control loops can optimize the data traffic and achieve effective utilization of the network bandwidth. It can be deduced that the good control performance can be achieved under proper selecting of MALD and optimal sampling time to the dynamic of the system.