GA-Based Control of a Robot Manipulator in a Foundry Workcell

In this paper, a method is presented to simulate the performance of an industrial robot manipulator in a foundry workcell. The robot under consideration is IRB 6640 180-2.55 from ABB used for handling heavy cores in ElectroSteel Castings Ltd., Kolkata. The proposed method considers the time-optimal control of robot motion to simulate the actual behavior of the robot. A model-based control technique is applied to control the robot's end-effector motion. Genetic Algorithm (GA) is implemented to optimize the gains in the proposed control method. Points used for defining the required path are taken from the actual program running in the workcell. Parameters needed to construct robot kinematic and dynamic models and constraints on robot motion are extracted from robot manuals and simulations on the same robot in ABB RobotStudio (R) environment. Modeling, simulation, optimization, and control are done in MATLAB (R)/SIMULINK (R).