Brushless wound-field synchronous machines (WFSMs) have long been employed in a wide range of applications due to their maintenance-free operation and high reliability. Brushless excitation is typically achieved through the deliberate injection of harmonic or subharmonic currents into the stator winding using specialized converters and complex winding arrangements. Alternative topologies, such as embedded exciter-based synchronous machines, have also been explored, utilizing magnetically decoupled dual windings for excitation. In recent years, multiphase machines have emerged as attractive alternatives to conventional three-phase systems, offering advantages such as lower converter switch ratings, improved fault tolerance, and enhanced reliability. By exploiting additional degrees of freedom, multiphase machines enable independent control of multiple spatial harmonic fields, facilitating rotor excitation and power transfer in WFSMs. Through appropriate stator current control, dual magnetic fields with different pole pairs can be generated and regulated independently, resulting in higher torque density, improved slot and core utilization compared with their three-phase counterparts. Motivated by these advantages, this paper presents the analysis and design of a six-phase Brushless Multiphase Synchronous Generator (BMSG). A 2/4-pole configuration is adopted to demonstrate the generation of dual magnetic fields from a single six-phase stator winding. The operating principles and key performance characteristics of the proposed machine are validated through finite-element analysis (FEA) and experimentally verified using a laboratory-scale prototype.
Ultra-high-speed electric drives operating beyond 100 krpm are attracting increasing interest in compact turbomachinery and electrified auxiliary systems. Although synchronous reluctance machines represent an attractive permanentmagnet- free solution, their extension to this operating range is simultaneously constrained by rotor mechanical integrity, converter voltage capability, and the strong interaction between electromagnetic design, manufacturing accuracy, and drive operation. This paper presents an integrated system-level approach for the development and experimental validation of a 100 krpm-class synchronous reluctance drive based on the coordinated use of two complementary enabling technologies: a multi-material axially laminated anisotropic rotor manufactured by laser-directed energy deposition and an open-end winding dual-inverter architecture. The complete development process is presented, including additive-manufacturing technology selection, material characterization, rotor fabrication, electromagnetic and mechanical design, converter implementation, and multidisciplinary optimization. Two prototype drives were developed and experimentally investigated. The machine validation comprises flux-linkage mapping, apparent inductance evaluation, maximum torque-per-ampere assessment, load tests, and acoustic measurements, while the comparison between experiments and finite-element analysis highlights the influence of manufacturing-related phenomena, including the diffusion region generated between adjacent deposited materials. The drive validation further assesses the proposed converter architecture under very-high-speed operating conditions. The presented results demonstrate that the coordinated design of the synchronous reluctance machine, additive-manufacturing process, and power converter is the key enabler of next-generation ultra-high-speed permanent-magnet-free electric drives.
Modular Uninterruptible Power Supply (UPS) systems and islanded microgrids require reliable plug-and-play (hot-swap) capability to ensure uninterrupted operation in critical applications such as data centers and telecommunication infrastructures. Conventional pre-synchronization methods based on Second-Order Generalized Integrators (SOGI) effectively align the fundamental voltage components but inherently reject harmonic distortion caused by nonlinear loads. As a result, a voltage mismatch between the incoming inverter and the distorted Point of Common Coupling (PCC) may occur, leading to transient inrush currents during connection, potentially triggering protection mechanisms and compromising system reliability. This paper proposes a practical and software-only control strategy for hot-swap operation under nonlinear loading conditions. The method is based on a state-dependent harmonic pre-synchronization mechanism that temporarily injects the PCC harmonic content into the inverter voltage reference during the synchronization stage. By forcing the incoming inverter to replicate the distorted grid waveform prior to connection, the proposed approach minimizes the instantaneous voltage gradient across the contactor. Immediately after connection, the harmonic injection is disabled, restoring standard harmonic rejection and steady-state power quality. The proposed strategy requires no additional hardware or sensors and can be implemented in existing digital control platforms. Controller-hardware-in-the-loop results at four operating points demonstrate reductions of 42.4–60.1% in the connecting-inverter current peak. These results confirm the effectiveness and practical applicability of the method for improving hot-swap reliability in modular UPS systems and other critical power applications.
With increasing switching frequencies of voltage source converters, the probability of resonance excitation in drive applications increases. For rotating electric machines in resonance condition, high voltages can occur within the winding. Due to significant stress to the insulation system, such conditions have to be avoided. Consequently, the knowledge of series resonance positions is essential, which means that frequency dependent impedance curves of electric machines have to be calculated or at least measured prior to commissioning. The latter means that suitable measuring configurations have to be defined. In this paper, a simple linear model and selected measuring configurations for its parametrization are presented. Several sets of equations for the parametrization of the model are summarized. For two of them, conclusive experimental determined curves are given.
The escalating penetration of Distributed Energy Resources (DERs) presents significant operational challenges to Low-Voltage (LV) and Medium-Voltage (MV) Distribution Networks (DNs), particularly in terms of voltage stability and bidirectional power management. Multiport Converters (MPCs) are an efficient, unified solution to aggregate diverse resources and enhance system resilience. This paper proposes a multi-stage design of MPCs. An optimisation-based methodology is presented for optimally sizing the MPC terminals, minimising total economic cost while simultaneously optimising technical indicators such as annual line losses and voltage stability. This methodology is validated across two distinct utility use cases: a four-terminal MPC in an LV network, which successfully integrates PV generation and EV charging, and a two-terminal MPC in an MV network that effectively mitigates voltage imbalances between parallel feeders. Furthermore, the work reviews suitable MPC topologies for these grid applications and proposes a qualitative comparison technique implemented in a MV application. Finally, some control architectures for the proposed topologies are thoroughly verified through Control Hardware-in-the-Loop (CHIL) and experimental converter prototype testing.
Precise thermal models of permanent magnet synchronous motors (PMSMs) are required for several purposes, like mechanical design, thermal management, drive control, and integration into electric vehicles. Identifying a precise thermal model based on measurements is typically a time-consuming and complex task, and the result depends considerably on the measurements available. To simplify this task, the current paper develops a method for optimally designing an experimental cycle in favor of parameter identification. The aims are to maximize the information contained in the measurements, to make the parameter estimation accurate and robust, and to reduce the total cycle time. The method uses a tailored optimization problem and an existing gray-box model of the PMSM to find optimum input trajectories for the experiment. The objective function considers the A-optimality criterion based on the Fisher information matrix. The method is validated based on both computer simulations and test bench measurements. The obtained results demonstrate that the cycle designed using the proposed method leads to reduced parameter variance in simulation and improved model accuracy when evaluated on test data from experiments.
This paper proposes a method to jointly estimate the state of the grid-connected voltage source converter (VSC) with an LCL filter and the grid inductance based on non-linear sigma-point Kalman filter (SPKF). The estimated parameters and states are used to improve the performance of the grid-connected converter controlled with modulated model predictive control (MPC). The proposed SPKF-MPC control method does not require sensing the point of common coupling (PCC) voltage. Moreover, the paper proposes a novel method to enable the proposed SPKF-MPC to distinguish between the steady-state and low-voltage ride-through (LVRT) operation and to detect faults. Simulation results have shown good performance for the proposed SPKF-MPC during steady-state and disturbances. In addition, the proposed SPKF-MPC has demonstrated improved performance over conventional MPC, particularly in the case of grid inductance uncertainty/mismatch and weak grid conditions. Results have been validated via experimental setup, verifying the accuracy and robustness of the proposed control strategy.
Increasing penetration of Wind Turbines (WTs) degrades the power system small-signal stability due to the complex modal interaction and reduced system inertia, which makes tuning of traditional Power System Stabilizers (PSSs) ineffective. This paper proposes a novel multi-stage optimization framework that revitalizes existing PSS infrastructure by co-optimizing their placement and parameters in coordination with WT controllers. It is based on the Genetic Algorithm (GA) with three key enhancements: 1) proactive pre-tuning stage that mitigates the Open-Loop Modal Resonance (OLMR) as a design constraint to guarantee the baseline stability; 2) Coherency-based Population Initialization strategy, derived from quantitative mode shape analysis, to ensure accelerated convergence; and 3) H∞-inspired Objective Function that integrates PSS placement logic to achieve the maximum damping of critical modes with the minimum control effort. The proposed framework is validated by the IEEE 39-bus test system with up to 20% wind power penetrations, achieving up to 55% increase in the damping ratio of critical modes compared to conventional methods. Simulations under N-1 contingencies confirm its robustness and faster oscillation damping. The proposed framework can provide a robust, computationally efficient and practical solution to maintain stability of grids with significant wind power penetrations without requiring costly hardware upgrade.
Permanent magnet synchronous motors (PMSMs) are broadly used in diverse applications due to their inherent advantages. Open-circuit faults (OCFs) are among the major fault classifications in PMSMs, posing significant concerns due to their contribution to torque ripples, vibrations, and efficiency degradation. Therefore, accurate and real-time OCF diagnosis is essential for reliable operation and predictive maintenance practices. This underscores the importance of a robust diagnostic framework that enables early fault detection and localization, supports embedded integration, and requires no additional dedicated sensors. However, existing studies rarely address these requirements together. To overcome these limitations, this paper proposes a novel OCF diagnostic framework that fuses features derived from multiple strategies, including wavelet energy-based features, frequency-domain features extracted from current waveforms, and speed measurement data. The extracted feature vector is used as input to a lightweight deep neural network (DNN). The proposed approach enhances interpretability and enables seamless embedded integration compared to conventional raw-data-driven machine learning models. Additionally, an extended refinement layer is incorporated to enable integrated fault detection and classification for OCF while enhancing diagnostic transparency. The effectiveness of the proposed method is demonstrated through MATLAB/Simulink simulations using the PLECS Blockset and further validated in real-time with an RTBox based hardware-in-the-loop (HIL) setup using a C2000 launchpad. Furthermore, experimental validation is conducted using a domain-adaptation strategy based on transfer learning. Performance evaluation confirms diagnostic accuracy exceeding 99% across varying operating conditions. The validation process achieves fault detection within 22% of a fundamental electrical cycle, with fault localization occurring within 40% of an average, demonstrating the robustness and adaptability of the proposed method. A sensitivity analysis of the proposed algorithm's feature vector validates the effectiveness of HF features. Furthermore, the risk distribution matrix provides insights supporting informed maintenance decisions
The emerging adoption of wide-bandgap (WBG) semiconductor devices in motor-drive systems poses significant challenges to the stator winding dielectric insulation of electrical motors due to their high voltage slew rate. Such ultra-fast switching characteristics generate fast-rising voltage pulses with high $dv/dt$, which may produce high-frequency overvoltage transients at the motor terminals. The reflected overvoltages result in a non-uniform voltage distribution along the motor stator windings, imposing excessive electrical stress on the winding insulation. Therefore, a high-frequency (HF) model operating at the MHz scale, capable of accurately predicting reflected voltage stress across stator winding coils and turns, is essential for developing high-reliability motor-drive systems. This paper proposes an HF modeling framework that integrates finite element analysis (FEA) with a distributed parameter network to improve the prediction accuracy of reflected voltage distributions. The proposed HF model incorporates frequency-dependent material characteristics, winding configurations, and inter-winding coupling effects through systematic parameter extraction, while maintaining moderate implementation complexity. A 2-hp induction motor prototype has been custom-rewound to experimentally verify the proposed HF model under multiple cable lengths and voltage pulse rise times. Experimental results demonstrate the superior performance of the proposed modeling approach compared to conventional methods, achieving time-domain reflected voltage estimation errors below 4%, while maintaining strong agreement with measured impedance spectra across a wide frequency range.
Mobile robot platforms are increasingly deployed in modern industrial environments, and ensuring their efficient operation has become an important aspect of system design. Their highly dynamic power profile, characterized by regenerative braking, fast acceleration, and frequent cycling, creates significant challenges for providing a stable and efficient energy storage. To meet these demands, modern mobile robot applications tend to integrate supercapacitor energy storage systems, either standalone or as part of a hybrid energy storage system (HESS). The supercapacitor modules are often implemented in different series and parallel configurations of cells to meet the application requirements. However, cell-to cell variations in capacitance, ESR, and aging behavior in the module lead to voltage imbalance, reducing usable energy and further accelerating degradation. Thus, to prevent these effects and maintain uniform cell utilization over time, an efficient voltage-balancing device is required. This paper proposes an Active Resonant Voltage Balancer (ARVB) as a solution to balance voltages between series supercapacitor cells within a module of a hybrid energy storage system (HESS) for a mobile robot. Compared to other active balancing methods, the advantages of the ARVB solution are high efficiency, high power capability, and low complexity, making it a suitable choice for this application. The guidelines are provided for the design optimization of the ARVB, based on the specified worst-case voltage imbalance in the supercapacitor module. The selected design is implemented and experimentally verified on an equivalent supercapacitor module for a single ARVB fundamental block. The benefit of using ARVB in the supercapacitor module is demonstrated by the increased total energy stored in the module at the end of the CC charging cycle. It shows an increase in stored energy up to 34% at a charging current of Ich = 3[A], compared to the supercapacitor module without ARVB. Two energy efficiency metrics were evaluated in this work. The charging energy utilization efficiency was evaluated with respect to the ideal balancing case and represents the achievable improvement in usable stored energy after charging enabled by the balancing system. A maximum charging energy utilization efficiency of 85.5% was obtained at a charging current of Ich = 3[A]. In addition, the balancing energy efficiency was evaluated during idle-state equalization, typically used in the literature. The maximum efficiency of 95.7% is obtained for the initial voltage difference Δu12 = 150[mV].
Multiphase motor drives offer several advantages over their three-phase counterparts, including reduced power per phase, lower voltage or current levels while maintaining the same output power, and enhanced fault tolerance. Over the past three decades, numerous studies have investigated modulation techniques for multiphase converters, as the increased number of phases introduces both additional challenges and new degrees of freedom. Among these challenges are the reduced linear modulation range and the more complex low-order harmonic behavior of multiphase machines. This paper proposes a space vector modulation strategy composed of four coordinated techniques designed to cover the entire modulation range of a two-level nine-phase converter. In the overmodulation region, discontinuous modulation methods are employed to simultaneously reduce switching transitions and control harmonic components. Additionally, a machine configuration with three isolated neutral points, instead of a single neutral connection, is adopted to mitigate current harmonics. Simulation and experimental results validate the proposed approach at low switching frequencies, which are particularly relevant for high-power applications. The performance of the proposed strategy is also compared with other modulation technique reported in the literature.
Finite Control-Set Model Predictive Control (FCS-MPC) has been extensively applied to multilevel converters due to its simplicity, fast dynamics, and ability to handle multiple objectives. However, its main drawback is the exponential increase in computational load as the number of switching states grows, which limits its real-time applicability in Multilevel Converters (MCs) with many levels. This paper proposes a novel strategy, FCS-MPC with Geometrical Pre-Selection States (FCS-MPC-GPS), that significantly reduces the number of evaluated states while preserving current tracking and capacitor-voltage regulation performance. The method introduces a geometric redefinition of the αβ voltage-vector frame, replacing the conventional hexagonal representation with concentric circles whose radii correspond to the voltage magnitudes. This transformation enables a hierarchical reduced-search selection process that separates current tracking and capacitor-voltage balancing into sequential stages. The proposed approach is experimentally validated on a three-cell Four-Level Flying Capacitor (4L-FC) converter with 512 switching states, demonstrating a computational load reduction of nearly 98.55% while maintaining accurate current tracking, capacitor-voltage regulation, and fast dynamic response.
The capacitor current active damping (CCAD) method is widely used to dampen the LCL-filter resonance in grid-connected inverters due to its simple implementation. Alternatively, it can be replaced by the capacitor voltage active damping (CVAD) as a low-cost solution, utilizing the capacitor voltage of LCL-filter for synchronization and damping, avoiding an extra sensor. A critical part of CVAD is the differentiator that is used to estimate capacitor current, yet, differentiators bring challenges due to their sensitivity to noise. To address these challenges, this paper proposes a family of enhanced discrete-time differentiators - Lanczos, Super Lanczos 1 and Super Lanczos 2 - for CVAD. This class of discrete-time differentiators based on finite impulse response filters is discussed, with detailed explanations provided on how to obtain the coefficients of the differentiators. The system was implemented using Controller Hardware-in-the-Loop (C-HIL) and experimental results demonstrate the effectiveness of the proposed method in damping the LCL resonance in a single-phase grid-connected inverter under steady state, under reference step conditions, and in the presence of noise.
This article presents an enhanced droop-based grid-forming control method for inverter-based energy resources connected to the transmission grid. The main advantage of inverter control—its ability to control frequency and phase independently—over conventional synchronous machine behavior is emphasized. This additional degree of freedom is used to implement an intrinsically damped response to grid disturbances such as voltage steps, frequency deviations, and phase jumps. The proposed control is analyzed in the state-space for three representative operating scenarios: grid parallel operation, two weakly coupled power resources, and a two-area system with two resources per area. A design goal for the control gain is formulated and applied to the operating scenarios under consideration, using only characteristic per-unit quantities and time constants. It is shown that the damping coefficient at the boundary case of all real-valued poles mainly depends on the intrinsic parameters of the power resource, such as short-circuit impedance and filter time constants, and is fairly independent of other system parameters. Simulation results confirm the analytical design of the damping coefficient, using a two-area system as a benchmark case. In comparison to conventional power plants, the proposed grid-forming control with an optimized damping coefficient demonstrates a significantly improved damped response to grid events, such as load or generation changes or partial tie line tripping. Experimental results gained in a complex power system laboratory validate the well-damped behavior for typical grid events.
The integration of electric vehicles (EVs) into microgrids via vehicle-to-grid technology offers a significant opportunity to provide demand-side flexibility. This article presents an experimental analysis of two load demand flexibility (LDF) strategies: LDF-FLAT and LDF-ZERO, implemented within a real-world university microgrid at Sapienza University of Rome, featuring 12.4 kW of PV generation and 6.5 kWh of stationary storage. Experimental tests demonstrate the system's effectiveness in transforming building consumption into a dispatchable resource. Specifically, the LDF-FLAT algorithm successfully flattened consumption peaks from similar to 40 kW to a constant target of 15 kW. In the LDF-ZERO scenario, the system achieved near-total grid independence, reducing the grid dependence ratio to 1.8% under sunny conditions and 2.6% under cloudy conditions. The vehicle to grid power ratio reached 97.2%, proving the EV's capability to act as the primary flexibility resource. Furthermore, the control logic ensures battery health by limiting the depth of discharge to 80%, potentially extending the battery life by 300 cycles. These quantitative results validate the reliability and efficiency of the proposed management system in meeting distribution system operator flexibility requirements.
The rising use of renewable energy sources in power networks has put rotational inertia in jeopardy, which has led to instability and system degradation. Demand Response (DR), a key component of the electric power system's future dependability, is used in this research. Within a DR framework, a modified three-DOF tilt fractional order integral derivative controller with a proportional fractional order derivative 3DOF(TFID)-(FOPD) is created for frequency management of a deregulated identical hybrid two area power system. There are hydropower plants with all the physical constraints, conventional thermal power plants, renewable energy sources (wind, STP, and biogas), and electric vehicle assets in each location. The proposed approach for adaptively tuning the intended controller's coefficients of the hybrid test system is called Chaos Quasi-Opposition-based Sea-horse Optimization (CQOSHO). CQOSHO's superiority has also been evaluated against a number of other renowned optimization techniques. Furthermore, the test system's stability is assessed. The improved results show that, when compared to the existing work and under various conditions, such as system uncertainties, physical constraints, and large penetration of Renewable Energy Sources (RES), the CQOSHO-tuned proposed controller offers a magnificent increase in system frequency stability. Finally, the viability of the proposed technique with source and load intermittency is shown by an experimental evaluation using OPAL-RT OP4200.
Recent solutions for oil and gas production have driven companies to operate in increasingly harsh environments, necessitating the development of more complex subsea variable speed drives (VSDs). These systems now include driving permanent magnet synchronous machines (PMSMs) due to their potential advantages, such as higher efficiency. Ensuring stability in such new systems may be challenging when sensorless operation and very long cables are employed. This work presents an analytical model to assess the stability of PMSM-based VSDs under conventional scalar control as well as with a stability control algorithm to address potential instabilities. The subsea drive system is modeled using a physics-based approach, including a novel motor cable model employing distributed parameters, expressed in the d-q reference frame. A small-signal model is derived to evaluate stability across different operating conditions. Experimental validation is performed using a reduced-scale test bench to replicate the drive system. The study identified an instability range, while the proposed stabilization loop significantly increases the range of stable operation. An experimental setup was used to validate the model. The experimental results show that the stabilization loop effectively reduced torque variation-driven speed oscillations.
Guaranteeing frequency stability in low-inertia power systems requires converter-based renewable energy sources to provide rapid response under tight operational constraints. This article presents a twofold contribution to the coordinated fast frequency control of grid-connected hybrid renewable plants consisting of grid-forming battery energy storage and grid-following photovoltaic generation and electrolyzer loads. First, a model predictive control (MPC) strategy is developed to coordinate these heterogeneous assets. Extensive robustness evaluations across an uncertainty space demonstrate that the proposed MPC achieves a 48%-50% median improvement in tracking accuracy and a 33%-48% reduction in energy storage requirements compared to traditional proportional integral (PI)-based and decentralized strategies. Second, the MPC's performance is enhanced through a novel prediction model that formulates system states by the converter's output power instead of angle differences as in Tie line power flow-based models. Unlike the traditional approach, this model avoids the need for augmented disturbance states and achieves full observability by formulating load disturbances as process noise. The proposed model further reduces median tracking error by 18% and achieves a 9% median energy storage reduction in the robustness analysis. The performance of the controllers is evaluated through hardware-in-the-loop experiments for various contingencies. A variance-based global sensitivity analysis reveals that parametric uncertainty, specifically concerning grid inertia, grid impedance, and power ramp limitations, dominates performance impacts over structural model-plant mismatch. The results demonstrate that the adoption of MPC offers substantial benefits for hybrid renewable plant control, while the proposed prediction model further maximizes control performance.
Data-driven thermal models increasingly replace direct temperature sensing and purely analytical thermal modeling. Their workflow comprises data acquisition on sensor-equipped prototypes, parameter optimization (training), and cross-validation prior to deployment. While training and deployment are well studied, design of experiment (DoE) for data acquisition remains underexplored, although it largely determines model performance and must respect costly, limited test-bench time. This work demonstrates a practical closed-loop online excitation method that shapes the dataset distribution via model predictive control (MPC), using a symmetric Jensen–Shannon divergence (JSD) between a reference and the predicted distribution from a recursively updated plant model. Test-bench trials reveal sim-to-real challenges, notably the need for remote edge computing to handle intensive data processing and optimization, and sensitivity to initial-state errors, which slightly reduces accuracy compared with simulation.
