Abstract:
This paper investigates a real time distributed controller for a 3 DOF delta robot using low-cost educational simple microcontrollers. The parallel computing technique is used were the computational load is divided among several microcontrollers networked to each other to implement control methods. More specifically, the computation is distributed among four microcontrollers (MCU’s) which are connected to each other using CAN bus protocol. The main MCU is used to compute the control law. Each of the remaining
MCU’s is connected to one actuator and its attached encoder to form an Intelligent sensor-actuator system (ISAS). At each sample time, the ISAS broadcast a message using the CAN bus to the main MCU containing the information about the motor position. Then according to the control law, ISAS receives the corresponding controller value that has to be applied to the motor. All required periodic, aperiodic and sporadic tasks, were
implemented and will be handled by these MCU’s. Using this design, the computation time of control law can be minimized and implemented using ARDUINO microcontrollers. More, this method increases the flexibility of the system for additional equipment and control by adding more nodes to the network. The results show the applicability of the proposed distributed controller, it can track different types of control signals with acceleration up to 9.8 m/s^2 (1g).
