Computationally Efficient Model-Free Predictive Control of Zero-Sequence Current in Dual Inverter Fed Induction Motor

Abstract

A model-free predictive zero-sequence current (ZSC) control is presented in this article for the dual-inverter-fed open-end winding induction motor (OEWIM). The proposed method is performed in a cascaded finite set predictive control approach to decrease the computational burden. In the proposed method, a state observer estimates the ZSC and its disturbances. The estimated ZSC is suppressed by a disturbance rejecting control loop that produces the reference of common-mode voltage (CMV). As a benefit, the motor parameters have not appeared in this process, which improves the robustness of the whole control system. To reduce the computational burden, the model-free predictive control (MFPC) is implemented in two cascaded stages. First, the predictive algorithm is performed for the voltage vectors (VVs) with zero CMV to detect the optimum zone of the space vector locations. In the second stage, the proposed MFPC is implemented with the VVs of the selected zone, and the optimum VV is achieved. In this way, the predictive algorithm is iterated 13 times. On the contrary, the conventional predictive algorithm is iterated 27 times for the VVs of the dual inverter. So, the calculation of the proposed method is reduced almost 50%. This feature is achieved without neglecting any of the VVs. Therefore, the performance of the proposed method is not degraded by reducing the calculations. A comparative analysis is carried out through simulations and experimental tests. The results confirm that while the proposed MFPC method is computationally efficient, its performance and robustness are also improved.