Model Predictive Control of PMSM with Computed Torque for Servo Press

The most important feature of the servo press is flexible slide motion because the servo machine can track the trajectory smoothly and accurately. A conventional method is field-oriented control (FOC) based on the mini-closed loop of the electrical machine position. However, the moment of inertia varies across slide positions, and the servo press is a strongly nonlinear system. So linear control method utilizing kinematics can not follow the curve very well. Meanwhile, this conventional method also can not reduce the harm of backlash that must exist in the mechanism of the servo press. So in this study, the proposed method is adapted to the closed loop of the crank position. This method allows Lagrangian Approach to set up a dynamical model. The torque reference of the computed torque algorithm is implemented by model predictive current control. A simulation and experimental studies are coded on a dual-core DSP and tested in a 1600 kN servo press prototype. These results show that the proposed method is effective and has been analyzed to be better than the accuracy of the position and speed control of the conventional method in detail. The contribution of this paper is that the proposed method could improve the accuracy of crank position and speed in the servo press.