Rank Ordering Criteria Based Weighting Factor Evaluation in Model Predictive Torque Control of Five-phase Induction Motor Drive

Predictive torque control (PTC) is an emerging and promising technology in electric drives. The effective performance of the PTC drive lies in its simple control law wherein different constraints can be easily added with weighting factors. However only empirical approximation methods are used for the determination of weighting factor. This paper presents a ranking weight method to find the weighting factor in predictive torque control of five-phase induction motor (IM). The weighting factor value is based on the rank ordering of the criteria (such as torque, flux, xy-plane current harmonics, CMV etc.) considered in the cost function. Assigning weights based on the rank of the criteria is more reliable than randomly assigning weights to the criteria because the decision makers are more confident on the ranks of the criteria than weights. Rank order centroid (ROC) weight technique is presented and compared with conventional weighting technique. The viability of the concept is verified by numerical simulations and the results are validated through experimentation.