Fuzzy logic controlled direct frequency converters modulated by an expert knowledge-based space vector technique

This paper presents two expert knowledge-based modulated direct frequency converters (XDFC). One uses a direct approach to model the converter operation (D-XDFC), and the other a fictitious link modeling (F-XDFC), The proposed schemes merge fuzzy logic, expert systems and space vectors to control the converters. Their main feature is to achieve a unity AC-AC voltage gain. Hence, eliminating the need for coupling transformers and allowing XDFC driven motors to be rated at the system's nominal voltage. Output and input currents are simultaneously controlled. An expert space vector modulation (XSVM) is presented, that selects the next converter state using a set of rules based on a sound understanding of the DFC operation. Thus achieves a notorious processing time reduction compared to similar performance control algorithms. The converters' commutation frequency is kept below 850 Hz, which reduces the converter's losses, switches' stress and increases the power level the proposed XDFC may handle.