Inverters, as one of the key components of electrical systems, have experienced a great evolution in the last decade, and their performance improvement is a challenge even today, leading to many researches on topologies and control schemes. This study introduces a new multilevel converter topology which is able to supply bidirectional current loads. The proposed structure has fewer power electronic devices such as power switches, driver circuits, power diodes, and DC voltage sources and, can be designed in both symmetric and asymmetric structures. In order to increase the number of output levels and the proposed basic unit development, modular expansion or cascading methods can be used. This study demonstrates that the aforementioned methods have the best results in asymmetric and symmetric structures of the proposed topology, respectively. The comparison between the proposed converter and some previous topologies shows that it has better conditions with respect to the used semiconductor count, switching and conduction losses, and total blocking voltage. The operation and performance of the proposed multi‐level converter have been ascertained through simulations and verified experimentally for a single‐phase symmetric thirty‐one‐level inverter which shows the proposed converter's ability in smooth sinusoidal output voltage generation with minimum total harmonic distortion.

This paper deals with a model reference adaptive system (MRAS) estimator based on second-order sliding mode (SOSM) strategy for speed-sensorless direct torque control (DTC) of six-phase induction motor (6PIM). A compensated flux observer is proposed using well-known super-twisting algorithm (STA), which is inherently an SOSM technique. This observer is adopted as the reference model in the MRAS-based speed estimator to overcome the chattering problem of the classical sliding-mode techniques. The estimator is robust against parameter uncertainties and dc-offsets. In addition, a super-twisting algorithm-based second-order sliding-mode (ST-SOSM) controller is designed for speed loop in the DTC strategy to increase its robustness under applying external load disturbances. The speed control loop needs the load torque signal. For this purpose, a ST-SOSM load torque observer is proposed, where it provides a robust performance without the penalty of high chattering. Simulation and experimental results confirm the validity and the effectiveness of the proposed approaches.

Any mismatch between the real and tuned values of the stator resistance (R ₈ ) leads to a degraded performance of the direct torque control (DTC) strategy. The underestimation of R ₈ causes an error in the stator flux and torque estimations, thereby weakening the DTC performance; however, an overestimation of R ₈ can make the DTC unstable. The main objective of this paper is to present a novel estimator based on the model reference adaptive system (MRAS) for online estimation and tracking of R ₈ in the DTC of six-phase induction motors (6PIMs). Voltage and current vectors of 6PIM are mapped into α - β, z ₁ - z ₂ , and o ₁ - o ₂ subspaces, where only the α - β components are related to electromechanical energy conversion. The proposed technique uses additional degrees of freedom, i.e., the z ₁ - z ₂ subspace to estimate R9. The attained merits include simplicity, low computation burden, and low sensitivity to other parameters of the 6PIM. The adaptation law is derived from the Lyapunov stability theory and Popov hyperstability theory. The proposed technique is applied under a duty cycle control based DTC, where z ₁ - z ₂ currents possess reduced amplitudes in comparison with the conventional DTC. The validity of the proposed system is verified by simulation and experimental results.

In this paper, a flux search controller is proposed to improve the efficiency of direct torque control of a six-phase induction machine. The proposed flux search controller is based on adaptive gradient descent of motor flux value with fast response and easy implementation. A proper loss model of six-phase induction motor in conjunction with the proposed method has been presented. Stator frequency variation with respect to flux reduction is considered as a new aspect of fast flux search control. Reduction of the flux in the search control technique leads to an increase in stator frequency, and an error in core loss calculation occurs if the frequency variation is ignored. The proposed search controller greatly improves efficiency by reducing core loss as well as harmonics loss. The approach not only is easy to implement and adaptive with regard to parameter variations but also requires no additional hardware for practical implementation. The simulation and experimental results presented in this paper verify the effectiveness of the proposed method in improving efficiency of direct torque control of six-phase induction motor drive.

In this paper, a new structure for symmetric and cascade multilevel converter is presented. This structure requires the least power electronic components, gate driver circuits, power diode, and dc voltage source. The number of output levels is high. Reduction of the number of power electronic switches against voltage levels is an important factor in multilevel converters due to safety in switches, circuit size, cost, installation area, and complexity in control. To maximize the number of output level in the proposed optimized structure in order to attain different goals, such as reduction of stress on switches, the result of comparing the proposed structure and other structures present. The proposed method is robust against the switching faults in multilevel converter for dc source and the equipment. The simulation and experimental results presented in this paper show the performance of the proposed method.