Shore-to-ship (S2S) power supply systems provide an effective approach to achieving the decarbonization of maritime transport. These systems require the use of voltage source converters (VSCs) operating in grid-forming (GFM) mode to ensure stable microgrid operation during connection and islanded transitions. As a method to synchronize the GFM converter with the ship generators, the international standard IEC/IEEE 80005-2:2016 recommends the usage of droop values established by the ship prior to connection with the shore. However, various techniques exist for designing the power control loop (PCL), and the implementation of droop values differs among these methods. This paper presents a comparative analysis of a traditional droop controller, a virtual synchronous generator (VSG), specifically, a synchronverter, and a virtual oscillator controller (VOC), when the same droop values are applied to all. Simulation results demonstrate that the droop controller and the synchronverter yield nearly identical responses. In contrast, the VOC exhibits variable equivalent droop gains as a function of the output voltage, leading to distinct dynamic behavior.

Shore-to-ship (S2S) power supply systems play a key role in the decarbonization of the maritime transport. These systems require the use of voltage source converters (VSCs) operating in grid-forming (GFM) mode. GFM converters must handle significant load impedance variations, primarily resulting from fluctuations in the power consumption of docked vessels. Moreover, the power demand of ships such as cruises and container vessels can reach several MW or even tens of MW, further complicating the integration of S2S connections in ports. Therefore, the design of a robust and high-performance GFM controller is essential. This paper proposes a single-loop voltage controller, designed using state-space control theory, for the operation of GFM converters in S2S applications. The proposed controller achieves accurate voltage tracking, fast transient response, and complete disturbance rejection at both the positive-and negative-sequence components of the fundamental frequency, due to its zero output impedance. In addition, the controller exhibits a favorable sensitivity function and wide stability margin. These features enable compliance with international standards for S2S connections.

Multiphase permanent-magnet synchronous machines (PMSMs) with fractional-slot concentrated windings (FSCWs) are favored for many uninterruptible applications due to their high fault tolerance and torque density. Two different configurations are often adopted: single layer (SL) and double layer (DL). For SL, each slot is filled with one phase winding, while for DL, two phases can share the same slot. Accordingly, many research papers advise adopting SL in fault-tolerant PMSMs, rather than DL, because SL offers higher physical, thermal, and electromagnetic isolation between phases. However, the lower thermal isolation, as for DL, may enhance the heat transfer. This could be expected to reduce the hot-spot temperature for DL. Thus, the established preference of SL over DL for fault-tolerant PMSMs may be questioned, and further analysis is necessary. This paper compares the temperature distribution of six-phase PMSMs between SL and DL FSCWs under both healthy and faulty conditions, considering open-circuit and short-circuit (interturn) winding faults. For accurate results, finite-element analysis and computational fluid dynamics are employed for the loss calculation and thermal simulation, respectively.

The reduction of emissions, noise and fuel consumption costs that ships incur while berthing is driving investment in the electrification of port power supply systems. Current regulations, such as the IEC/IEEE 80005, require vessels to shut down their onboard generators and connect to shore power for as long as practicable during port stays. The method of supplying electrical energy to ships from shore power stations is known in the shipping industry as cold ironing. The shore-to-ship grid consists of a 60/50-Hz ac-voltage bus, rated from 1 kV to 15 kV for medium-voltage systems, and up to 1 kV for low-voltage systems. These grids can be supported by multiple grid-forming (GFM) power converters connected in parallel. These converters must interface with the ship’s power plant both when the onboard generators are running and after they are shut down. In such conditions, the converters often supply different types of loads and experience variable loading configurations depending on the type and size of the vessel. The parallel connection of GFM converters, combined with the variable impedance of the ship’s power plant can affect the stability of the shore-to-ship power connection. It is fundamental that each GFM converter sees at its output an impedance that is at least equal to the minimum impedance required to ensure stability. This article focuses ensuring stability in shore-to-ship connections through a comprehensive design of the voltage and droop controllers, in combination with virtual impedance (VI). The VI helps prevent instabilities when GFM converters are connected in parallel, regardless of the impedance seen from their output terminals.

Six-phase permanent-magnet synchronous machines (PMSMs) offer multiple benefits compared with three-phase ones, e.g., increased torque density and reduced phase-current rating. A method has been recently proposed to generate current references ensuring minimum stator copper loss (SCL) per torque for nonsalient PMSM drives with sinusoidal or nonsinusoidal back-electromotive force (back-EMF), while the peak currents were suitably limited (with harmonics) to the maximum instantaneous currents of the inverter even during transient overloads. Furthermore, even greater mean torque could be attained by permitting a certain (limited) torque ripple when needed. However, such technique did not take into account the voltage limits imposed by the dc-link voltage. This paper proposes an alternative method based on look-up tables (LUTs) that provides these functionalities for symmetrical six-phase nonsalient PMSMs, while effectively considering the voltage constraints in the optimization. This makes it possible to minimize the losses and torque ripple per mean torque in not only the entire torque range, but also in the full speed range. To decrease the computational requirements, the offline LUT-generation process is designed so that it identifies the range of operating points where the solution changes, and then the optimization problem is solved only within such range.

Recently, it was proposed to combine circulating-current (x-y) filters with a carrier-based pulsewidth-modulation (PWM) method, which injects harmonics in no-torque x-y subspaces (attenuated by the filter) of five-phase motors to achieve high dc-link utilization (DLU). The filter allows prolonged high DLU without overheating. Unlike square-wave switching, no torque ripple is introduced by the PWM. Compared with other PWM techniques capable of high-DLU operation (often called overmodulation), the one proposed for said filters is especially convenient because of its simplicity. However, it is unclear if this PWM method can be extended to machines of any phase number n and winding arrangement (symmetrical or asymmetrical). This paper extends this simple carrier-based PWM strategy to n-phase (symmetrical or asymmetrical) drives, and evaluates its behavior in terms of DLU. The improvement in DLU, compared with conventional PWM with zero-sequence injection, is considerable in most cases. Current distortion is negligible with suitable filter, and no torque ripple (unlike square-wave switching) is generated.

In this work, a time delay estimation based optimal conventional and second order sliding mode stator currents controller are proposed and compared for a six-phase induction machine fed by voltage source power converters. In a first step, the speed is regulated in an outer loop using a proportional-integral controller. Then, in a second step, the stator currents are controlled in an inner loop using the proposed methods. On one hand, the first method is a combination of time delay estimation with optimal first order sliding mode with exponential reaching law. On the other hand, the second method is a combination of time delay estimation with optimal super-twisting control. Based on Lyapunov theory, the stability of the stator currents closed-loop error dynamics is demonstrated and sufficient conditions of stability are determinate. As a case example, numerical simulations have been conducted on an asymmetrical six-phase induction machine to demonstrate the good features of the developed nonlinear controllers.

This paper investigates the problem of stator current control in presence of uncertainties and unmeasurable rotor current for a six-phase induction motor drive. An inner control loop based on a robust discrete-time sliding mode with time delay estimation method is proposed to ensure the finite-time convergence of the stator currents to their desired references while the proportional-integral controller is used for the outer speed control. Sufficient conditions are established to ensure the stability of the closed-loop system. Simulation results were carried out to verify the performance of the proposed robust control strategy for a six-phase induction motor drive.

The current control has an important impact on the performance of three-phase voltage source converters. A common approach is to adopt a two-degree-of-freedom scheme by which the reference-tracking and disturbance-rejection response are independently adjusted. On the one hand, a conventional current controller, such as a proportional-integral or resonant controller, is tuned to optimize the reference-tracking dynamics. On the other hand, an active resistance is selected to improve the disturbance rejection without altering the reference tracking. There are methods to further enhance the disturbance rejection by, e.g., adding extra selective controllers, but this increases complexity and requires the knowledge of the frequencies to be compensated. Therefore, it is still desirable to ameliorate the disturbance rejection to a greater extent, in a simple manner and without a priori information about the disturbance frequencies. This paper proposes to improve the disturbance rejection by means of an active inductance, without affecting the optimum reference tracking. Concerning the transient response, the overshoot (and hence, the integrated error) is significantly reduced. Furthermore, the steady-state attenuation of harmonics becomes larger in the lower part of the spectrum, where the harmonic disturbances usually arise. Simulation results supporting the functionality of the proposal are provided.

Multiphase machines are attracting significant attention due to their advantages over three-phase ones. When using pulsewidth modulation (PWM), a zero-sequence signal can be injected so that the dc-link utilization is improved. For instance, in five-phase machines, for a certain dc-link voltage the fundamental component is increased in this way in 5.1 %. Instead of PWM, it is possible to apply a square voltage waveform, which permits even better dc-link exploitation (27.3% with respect to conventional without zero sequence). Given the substantial voltage distortion, filters with large impedance at the low-impedance no-torque subspaces were devised. In any case, some of the square-waveform harmonics are mapped into the torque-producing plane without being attenuated by the filters, thus generating torque ripple. This paper proposes a carrier-based PWM method for five-phase drives that include a filter for circulating currents. This PWM technique takes advantage of the no-torque-producing components, which see a high filter impedance, to exploit the dc link to a degree similar to square voltage waveforms (23.1 % versus 27.3%), but without the extra torque ripple. Simulation results are provided.

Multiphase drives are particularly attractive for safety-critical applications, due to their enhanced fault tolerance. Most of the faults in electric drives occur in the power electronics. Switch faults can be divided into open- and short-circuit ones. The conventional techniques to handle switch faults in two-level converters mainly consist in either keeping the corresponding phases open or connecting the stator terminal to the de-link midpoint through a triode for ac (TRIAC). However, the former results in lower maximum torque and larger stator copper loss (SCL), whereas the latter leads to a reduction in the maximum speed and has not been considered in the context of two-level multiphase drives. This paper proposes to adequately combine these approaches in two-level multiphase drives, depending on the speed, the faulted switches and the neutral configuration, so that the total operating region of the drive is extended and the SCL is decreased, when compared to the conventional strategies. It is also shown and exploited that, for multi phase drives based on multiple three-phase winding sets, turning on or off a TRIAC placed between their neutrals as a function of the speed allows further improvement of the postfault performance. Simulation results are provided to confirm the theory.

Synchronization techniques can be broadly classified into two major categories-closed-loop and open-loop methods. The open-loop synchronization (OLS) techniques, contrary to the closed-loop ones, are unconditionally stable and benefit from a fast dynamic response. Their performance, however, tends to worsen in the presence of frequency drifts. To deal with this problem, two approaches are often recommended in the literature-adapting OLS techniques to the grid frequency variations by feeding back the frequency estimated by them or using the frequency estimated by a secondary frequency detector in a parallel manner. In the presence of the frequency feedback loop, nevertheless, the OLS technique may not be truly open-loop, which makes a deep stability analysis necessary. Using the secondary frequency detector, on the other hand, increases the computational effort and implementation complexity. Another drawback of most of the available OLS techniques is that their implementation involves the computation of sine and cosine functions, which is undesirable from the computational standpoint, particularly when the implementation with low-cost digital signal processors is intended. The aim of this paper is to develop a true OLS (and therefore, unconditionally stable) technique without any need for the calculation of sine and cosine functions. The effectiveness of the proposed synchronization technique is confirmed through the simulation and experimental results.

The time delay in the current control loop of a grid-connected voltage-source converter (VSC) may cause instability if the VSC is equipped with an inductive-capacitive-inductive (LCL) input filter. However, instability is prevented if the input admittance of the VSC can be made passive. This paper presents a systematic method for gaining passivity. Thereby, stability is obtained for any LCL-filter resonance and for any passive grid impedance. The method is equally applicable to single-phase and three-phase systems.

The multi-motor drive concept is based on the use of the additional degrees of freedom available in multiphase motors to independently control a group of series-connected machines fed from a single voltage source inverter. To achieve this independent control, the motors should have a perfectly sinusoidal distribution of the stator windings. However, due to non-linearities in the real motors, low order harmonics appear in the system. Depending on their origin, two main types of harmonics are present in multiphase machines, time and spatial harmonics. These harmonics produce additional losses and parasitic torques in the series-connected machines. This paper analyses the relationship between low-order spatial and time harmonics and its effects in a series-connected five-phase two-motor drive through experimental results collected from a laboratory setup.

Low-order odd harmonics arise in practical multiphase drives because of the machine and converter nonlinear behavior (e.g., deadtime and saturation). If the windings are distributed, some of the harmonics cause torque ripple, while others cause losses. The latter is aggravated by the small impedance in the no-torque subspaces. The harmonics can be compensated without steady-state error by proportional-integral controllers in multiple synchronous reference frames (SRFs); however, a heavy computational load is required. In three-phase systems, the computational burden of this multiple SRF (MRF) scheme is often avoided by implementing instead resonant controllers (RCs) tuned at the harmonics that are multiples of six in an SRF rotating with the fundamental frequency. A similar structure has been proved to be satisfactory for deadtime compensation in an asymmetrical six-phase machine. Part 1 of this paper extends that harmonic current control strategy to symmetrical machines of any phase number. The optimum frequencies for the SRFs and for the RCs in each plane, so that the total number of RCs is minimized, are assessed. Part 2 studies the computational load and compares it with that of the MRF scheme.

It has been previously suggested to use multiple reference frame (MRF) structures, based on proportional-integral controllers in synchronous reference frames (SRFs), to suppress current harmonics caused by nonlinearities in multiphase drives. However, the rotational transformations are computationally demanding. In Part 1, the three-phase strategy based on resonant controllers (RCs) implemented in an SRF has been extended to any phase number. Nevertheless, while the computational burden of MRF and multiple RC (MRC) schemes has been assessed for three-phase systems in previous papers, it is yet to be studied for other numbers of phases. Part 2 analyzes the computational load of the MRC and MRF strategies as a function of the phase number and of the highest harmonic order to be compensated. It is proved that the former provides a resource saving with respect to the latter that decreases with the phase number and increases with the highest harmonic order under consideration. It is also concluded that the MRC strategy is particularly preferable in machines with even phase number (e.g., a multiple of six).

Multilevel multiphase inverters are a promising practical solution for high-power applications. In applications where high switching frequency cannot be accommodated due to the increased switching losses associated to the pulse-width modulation techniques, the space vector control modulation technique is a feasible alternative. The space vector control reduces the number of switchings by selecting one single switching vector during the whole switching period. This paper presents a multiphase space vector control modulation technique based on a multilevel multiphase space vector pulse-width modulation that can be implemented in real-time in a low-cost device. The five-phase case of the proposed technique is validated by simulation, and it is tested with a nine-level inverter feeding a five-phase induction motor.

The torque and flux of sinusoidally distributed multiphase machines can be controlled by using only two stator d-q components. Thus, in these machines there exist additional degrees of freedom, which in multimotor drive systems are used to independently control a group of series-connected multiphase machines. Series connected machines share stator currents, therefore it is interesting to study how harmonics produced by non-linearities, faults or imbalance of the feeding system are distributed in the two-motor system, in terms of subspaces and sequences. This paper analyzes low-order harmonic distribution and its effects in a series-connected six-phase two-motor drive through experimental results collected from a laboratory setup.

Grid codes force generation units based on renewable sources to act similarly to conventional power plants: when a fault is detected, they must remain connected and support the grid. To fulfill these requirements, the transient response of the current controllers is critical. This paper applies a recently proposed methodology which assesses and optimizes the transient response of proportional-resonant controllers (PR) to a different type of linear regulators: the vector proportional-integral (VPI) controllers. This methodology is based on the study of the error signal roots: both reference tracking and disturbance rejection abilities should be considered for proper gain tuning. This work proves that VPI controllers lead to longer settling times and bigger overshoot than PR controllers for high sampling frequencies. However, as the sampling frequency decreases, the transient response of both controllers presents more similarities. A three-phase voltage source converter prototype has been implemented. Experimental results provide a direct comparison between the transient behavior of VPI and PR controllers in different conditions: a +90° phase-angle jump in the current reference and a `type C' voltage sag at the point of common coupling.

The design of synchronous multiphase DC-DC converters for automobile applications is now a very active field, because the automotive industry forecast that future power demands inside a car will oscillate between 2.5 kW and 3.5kW, keeping a dual system of 42/14V batteries. The design of controllers for the optimal behavior of such converters is a very delicate task. In this paper, an optimized fuzzy control algorithm has been developed to control a synchronous multiphase converter of 1.6kW. First, the fuzzy control algorithm is designed and verified, together with a non linear model of the converter power stage, by means of Matlab and Simulink. Then, the fuzzy controller hardware is developed through Xilinx System Generator for Simulink, and implemented in a Spartan 3 FPGA to achieve a real-time controller