This study explores alternative framework configurations for adapting thermal machine learning models for power converters by combining transfer learning (TL) and federated learning (FL) in a piecewise manner. This approach inherently addresses challenges such as varying operating conditions, data sharing limitations, and security implications. The framework starts with a base model that is incrementally adapted by multiple clients via adapting three state-of-the-art domain adaptation techniques: fine-tuning, transfer component analysis, and deep domain adaptation. The Flower framework is employed for FL, using federated averaging for aggregation. Validation with field data demonstrates that fine-tuning offers a straightforward TL approach with high accuracy, making it suitable for practical applications. Benchmarking results reveal a comprehensive comparison of these methods, showcasing their respective strengths and weaknesses when applied in different scenarios. Locally hosted FL enhances performance when data aggregation is not feasible, whereas cloud-based FL becomes more practical with a significant increase in the number of clients, addressing scalability and connectivity challenges.

DC-link current harmonics and common mode voltage (CMV) are key design challenges for inverter-based power electronic systems. Addressing them collectively without additional hardware and/or complexity has promising advantages. This article investigates the interleaved utilization of modified space vector PWM schemes for parallel-inverter systems, targeting a combined and simultaneous reduction of these quality concerns. A time-domain analytical dc-link current formulation and a new application of sequence-based interleaving are combined in an offline numerical optimization algorithm that locates optimal interleaving shifts synced with the PWM sequence. With simulation and experimental validation, the corresponding numerical results ascertain an effective suppression of dc-link current harmonics alongside CMV reduction. In addition, this article extends the proposed idea to a special application for 3x multiphase machines through experimental validation.

This article proposes a simplified discrete space vector modulation (DSVM)-based model predictive direct speed control (MPDSC) with an improved load disturbance observer for permanent magnet synchronous motor (PMSM) drives. First, a simplified DSVM method is used to improve the steady-state performance of MPDSC. In this DSVM method, a novel geometric method relying only on three auxiliary lines in each sector is designed to simplify the algorithm’s complexity. In this way, the set of candidate vectors is quickly determined. Then, the current pulsation and speed of MPDSC are suppressed, and the computational burden of the DSVM execution process is reduced. Second, the reasons that affect the dynamic performance of the conventional linear extended state observer (ESO)-based mechanical disturbance observer are analyzed, and the observed error of the observer is derived. Based on the observer error, an improved mechanical disturbance observer is proposed to accelerate the convergence process. The Lyapunov theory proves the stability of the proposed observer. Finally, the feasibility of the proposed method is verified by experiments.

A multiphase induction machine model using vector space decomposition provides insights into many space harmonics through decoupled reference frames. In order to utilize this potential of the multiphase machine, the parameters of each vector space must be identified. These parameters are usually identified using standard no-load and locked rotor tests of each torque-producing vector space. However, these tests do not provide a method for identifying stator leakage inductances, which have comparable magnitudes to the magnetizing inductances for higher-order vector spaces. Thus, the magnetizing current cannot be neglected during the locked-rotor test. This paper proposes an accurate method for identifying the T-model parameters of a multiphase induction machine, which can be conveniently translated into the model of choice. The proposed method use harmonic relations to segregate the stator leakage inductance from the total stator inductance, enabling accurate locked-rotor and partial-load tests for the measurement of rotor side parameters. Lastly, a 9-phase induction machine is used for the evaluation of the proposed parameter measurement method. These measured parameters are used for the field-oriented control of each torque-producing vector space, whose performances are analyzed by quantifying the measured output torque error and linearity.

This article introduces a sensorless control strategy for a seven-phase induction motor drive, leveraging the inherent degrees of freedom in multiphase systems. The proposed approach relies on injecting a low-frequency current to excite the third spatial harmonic of the magnetic field in the air gap. This excitation facilitates the estimation of the third spatial harmonic component of the flux linkage. The injection of a low-frequency current enables the estimation of the rotor flux position at zero and very low speeds without torque ripple, which is an essential requirement for the reliable operation of sensorless drives. The experimental tests conducted under different operating conditions confirm the robustness of the proposed solution, as well as its dynamic response and stable behavior at zero and low speeds.

This paper investigates the impact of switching frequency and modulation technique on the efficiency of electric drivetrains utilizing inverter-driven Permanent Magnet Synchronous Machines, focusing on heavy-vehicle applications. Accounting for inverter conduction and switching losses, along with machine copper, iron, and magnet losses, a comparison between two- and three-level inverters is performed at key operating points relevant to long-haul truck applications. This study employs current-based finite element analysis for machine loss estimation and a virtual prototyping method for inverter loss estimation to analyze drivetrain efficiency. The findings indicate that there is an optimal switching frequency evidently improving overall efficiency. Moreover, the system using the three-level inverter demonstrates notably lower system losses.

Induction machine-based multiphase electrical machines combine the robustness and reliability of well-known induction machines with the benefits of additional degrees of freedom, such as potential higher torque density and true fault tolerance. The subcategory of variable phase-pole machines allows extending the operational torque and speed range with the same voltage and current limits by changing the number of magnetic poles through control without hardware reconfiguration. Maintaining the torque throughout the operation is vital for many applications. The question of maintaining torque during a pole transition still needs to be answered, especially concerning the electric limitations. This paper's proposed pole transition strategy minimizes the torque dip during the pole transition. Field weakening is added to the control scheme to accommodate the voltage limitation. Experimentally controlled, loaded, and scheduled pole transitions demonstrate the capabilities of the proposed strategy with and without voltage limitation.

Conventional model predictive current control of permanent magnet synchronous machines (PMSMs) relies heavily on a precise mathematical model, which may be challenging to obtain in certain cases. To address this issue, this article proposes a model-free predictive current controller for PMSMs. Specifically, the dynamic model of the motor currents is represented using two ultralocal models, which express the derivatives of a controlled output as the sum of an amplified control input and an unknown offset term. All system parameters, nonlinear terms, and unmodeled dynamics are encapsulated into two offset terms. Online estimation of these two offset terms is performed using a second-order fixed-time convergence observer without requiring exact knowledge of the system parameters in advance. In contrast to both the linear extended state observer and the super-twisting observer, the fixed-time observer enhances convergence speed while concurrently sustaining minimal chatting. To get rid of the need for model rated parameters, an extremum-seeking approach is utilized to tune the control gains in the ultralocal model, where the amplitude of the disturbance item in extremum-seeking approach is adaptively attenuated, resulting in effective reduction of the fluctuations of both the control gains and PMSM's currents in steady state. Subsequently, a predictive controller is designed using this ultralocal model. Finally, the effectiveness and advantages of the proposed control scheme are confirmed through experimental results.

Variable phase-pole machines possess the capability to expand their operational range by employing magnetic pole change while adhering to the same electrical voltage and current limits. This article proposes a strategy to achieve a pole change with constant torque while simultaneously optimizing current distributions for stator copper loss minimization. The overall outcome is the ability to facilitate smooth $dq0$ current transitions while reducing the energy consumption of variable phase-pole machine drives. Experimental results affirm the efficacy of this approach during the pole change process.

A multiphase induction machine (MPIM) modeled using vector space decomposition (VSD) provides insight into space harmonics which are segregated into vector spaces. Depending on the winding configuration of the MPIM, some vector spaces can produce torque in addition to the fundamental vector space, and they can be represented by a standard equivalent circuit (EC) of an induction machine (IM). This paper evaluates multiple EC parameters identification methods (IDMs) and proposes a robust IDM routine for their accurate estimation. Furthermore, the proposed IDM routine shows reduced sensitivity to errors in known parameters compared to the standard IDMs, which is validated with the help of a detailed simulation model of a 9-phase IM.