Search

Annual Energy Conversion Congress and Exposition, ECCE (11th. 2019. Baltimore, USA), This work was supported in part by the Research, Technological Development and Innovation Programs of the Spanish Ministry Economy and Competitiveness, under grant MINECO-17-ENE2016-80047-R and by the Government of Asturias under project IDI/2018/000188 and FEDER funds

IEEE Energy Conversion Congress and Exposition, ECCE (13th. 2021. Virtual), This work was supported in part by the Research, Technological Development and Innovation Programs of the Spanish Ministry of Science and Innovation, under grant PID2019-106057RB-I00.

IEEE Energy Conversion Congress and Exposition, ECCE (13th. 2021. Virtual), This work was supported in part by the Research, Technological Development and Innovation Programs of the Spanish Ministry of Science and Innovation, under grant PID2019-106057RB-I00.

This paper discusses the benefits of using silicon carbide (SiC) devices in a three-stage modular power electronic transformer. According to the requirements to be fulfilled by each stage, the second one (the DC/DC isolation converter) presents the most estimable improvements to be gained from the use of SiC devices. Therefore, this paper is focused on this second stage, implemented with a SiC-based dual active bridge. Selection of the SiC devices is detailed tackling the efficiency improvement which can be obtained when they are co-packed with SiC antiparallel Schottky diodes in addition to their intrinsic body diode. This efficiency improvement is dependent on the dual active bridge operation point. Hence, a simple device loss model is presented to assess the efficiency improvement and understand the reasons for this dependence. Experimental results from a 5-kW Dual Active Bridge prototype have been obtained to validate the model. The dual active bridge converter is also tested as part of the full PET module operating at rated power.

IEEE Energy Conversion Congress and Exposition, ECCE 2020 (12th. 2020. Detroit, USA), This work was supported in part by the Research, Technological Development and Innovation Programs of the Spanish Ministry Economy and Competitiveness, under grant MINECO-17-ENE2016-80047-R and by the Government of Asturias under project IDI/2018/000188 and FEDER funds.

Small grid-tied wind turbines based on permanent magnet generators often use a cost-effective power converter topology consisting of a passive rectifier, a boost converter, and an H-bridge inverter. Speed or position sensors are rarely used due to cost issues. Model-based estimators relying on electrical magnitudes are used instead. However, such estimators are parameter sensitive, which limit their accuracy. Further concerns arise if these parameters change with the operating condition of the machine, mainly due to temperature. Speed sensorless control using the rectifier voltage ripple is analyzed in this paper. This technique provides good dynamic response and does not depend on machine parameters. Simulations are provided for speed and power tracking comparison with an accurate model-based speed estimation method operating at non-rated parameters. They show the speed accuracy and power tracking capability of the proposed method are similar to that provided by a speed sensor. This is translated into a 0.9% power increase when the model-based speed estimator shows 9% of error. Experimental results are carried out to test the effect of current and temperature in the estimation, showing temperature insensitivity and some distortion due to fast current transients. A speed estimation accuracy of zero mean error and 1.7% standard error is experimentally obtained in the regular operation of the wind turbine.

Torque production capability of permanent magnet synchronous machines (PMSMs) depends on the magnetization state of the permanent magnets (PMs). Electrical stress, thermal stress, or the combined effect of both can produce demagnetization of the PMs, which eventually can result in several adverse effects including decrease in the motor torque and efficiency and increase of the torque ripple and vibration, eventually degrading the performance and reliability of the motor and drive system. A number of approaches have been proposed for detecting PM demagnetization using model-based flux estimation, signal injection, and spectrum analysis of current/back electromotive force; however, all these methods show limitations in terms of invasiveness, implementation cost, and/or reliability of the diagnosis. In this paper, accurate PM demagnetization detection based on measurements from Hall-effect sensors is proposed. Such sensors are often mounted in commercial PMSMs; the proposed method can, therefore, be implemented at practically no cost.

Control of permanent magnet synchronous machines (PMSMs) requires rotor position measurement/estimation, as well as the magnet polarity detection for startup of the machine. The rotor position of PMSMs is typically measured using a speed/position sensor (e.g., an encoder), while the magnet polarity is commonly measured using digital hall-effect sensors. Since simplifying the system can help improve reliability and reduce the cost associated with the measurements, the use of low cost analog hall-effect sensors for the estimation of the PM position, and polarity for control of PMSMs drives is evaluated in this paper. Experimental results performed on a 7.5-kW motor confirm that analog hall-effect sensors can replace both speed/position sensors and digital hall-effect sensors while maintaining adequate performance of the drive at a lower cost.

This paper analyzes the use of the voltage distortions in pulse width modulation voltage-source inverters caused by the inverter switching for islanding detection purposes. The nonideal characteristics of the inverters, mainly due to the dead time needed to have safe commutations, produce fundamental frequency-dependent harmonics (−5th, 7th, …) in the output voltage. These harmonic order are, in principle, an unwanted effect, as they reduce the power quality. However, they can potentially be used for islanding detection purposes. The physical principles of the method would be the same as for high-frequency signal injection methods that have already been proposed, but without the need of injecting a high-frequency signal, behaving, therefore, as a passive islanding detection technique.

Knowledge of the permanent magnet (PM) temperature in PM synchronous machines (PMSMs) is of great importance both for control and monitoring purposes. Increase in PM temperature during motor operation can degrade the magnetic flux strength and consequently the machine's torque production capability, and can also cause irreversible demagnetization of the PM. Direct measurement of the PM temperature is not viable in practice due to both cost and reliability issues. Indirect PM temperature estimation methods recently studied require knowledge of thermal or electrical model parameters or can have undesired effects on motor operation. In this paper, the feasibility of using low-cost Hall-effect sensors for PM temperature estimation is investigated. Hall sensors are present for detecting the initial position of the rotor in majority of PMSM applications for which incremental encoders are used for control. The proposed method can, therefore, be implemented with low or no additional cost. Experimental results on two interior PMSMs show that the method is capable of providing noninvasive estimation of the PM temperature without a priori motor parameter information for monitoring and protection against excessive increase in temperature.

Soft-stall control of small wind turbines is a method to protect the generation system or the load from high wind speeds while keeping the system generating. Soft-stall activation must be due to an excess of power or torque/current. This paper proposes a method to improve the existing soft-stall methods for over torque/current protection by the development of a turbine torque estimator. Due to the inconvenience of the continuously changing weather conditions, wind turbine emulators are broadly used by researchers instead of actual wind turbines. This paper also proposes two methods to emulate the wind turbine inertia without communications between the load drive (wind turbine emulator) and the generation system controller.

Thermal monitoring is a common feature in most electric machine drives since thermal overloading is one of the most common causes of motor failures. Contact-type sensors are normally used to measure the stator temperature in electric machines. However, use of this type of sensor is not advisable in the rotor as it requires cabling to a rotating part or the use of a wireless transmission system. Consequently, measurement of the rotor temperature is not easy in practice and is not normally implemented in standard machines. An alternative to rotor temperature measurement is rotor temperature estimation. To date, only thermal models have been used for rotor temperature estimation in doubly-fed induction machines (DFIMs). This paper proposes rotor temperature estimation in DFIMs using high-frequency signal injection. The proposed method estimates the rotor temperature from the rotor high frequency resistance, which is a function of the rotor windings temperature. The method does not interfere with the normal operation of the drive and can be implemented in existing DFIM drives without requiring additional hardware.

In a solid state transformer (SST), one of the most important roles is played by the dc-dc converter interfacing the high voltage port and the low voltage port while providing galvanic isolation. This paper proposes a quad-active-bridge (QAB) converter as the basic cell of a Modular Multilevel Converter (MMC) based solid state transformer for the integration of distributed energy resources (DER) and distributed energy storage systems (DESS) into the grid. This study comprises power transfer analysis, modeling and control of the QAB for this application. Furthermore, this paper shows an alternative method based on the star-mesh transformation to built the multiwinding tranformer equivalent circuit, that relates any two ports of the multiple-active-bridge (MAB) converter through a link inductance, which allows to describe power flow. Results are obtained by simulations in PLECS.

This paper addresses the design and implementation of distributed control structures for Modular Power Electronic Transformers (PETs). Medium voltage modular PETs are characterized by a large count of controlled power devices (easily in the range of hundreds or even thousands) and strict isolation requirements, both between modules in the high-voltage side and between the primary and secondary sides of each module as well. Due to this, distributed control structures have significant advantages compared to centralized ones. However, design of the distributed control structure is not trivial. Energization and start-up of the PET, as well as the required reconfigurations in the event of a module failure, to maintain the PET operative, can be especially challenging in this case. This paper discusses potential flexible, fault tolerant distributed control structures for modular PETs. Performance of the proposed solutions is verified experimentally.

This paper analyzes a cost-effective modification of the power topology commonly found in small wind turbine systems based on a passive rectifier and a boost converter. The boost-converter inductor and the input filter capacitor often placed at the rectifier output can be replaced by the generator-phase inductance. Different controller structures have been proposed for this low-cost inverter, but they have been focused on the converter itself rather than the overall turbine control. Moreover, only steady-state behavior has been demonstrated. This paper proposes a control structure only requiring retuning of the boost-current controller found in systems equipped with boost inductance, with other control loops remaining unchanged. The inductorless converter dynamic performance is studied and compared with the conventional topology in terms of current- and torque-control capability. The system efficiency, including the losses’ distribution in the generator, is analyzed. Simulation and experimental results are presented to demonstrate the technical viability of this proposal.

International Symposium on Power Electronics (Ee) (20th. 2019. Novi Sad, Serbia), The research was performed with the support of the Russian Science Foundation grant (project 15-19-20057).

The injection of a high-frequency signal in the stator via inverter has been proven to be a viable option to estimate the magnet temperature in permanent magnet synchronous machines (PMSMs). Injection of a pulsating d -axis current is especially appealing, as it is insensitive to the speed, and $L_{d}$ and $L_{q}$ inductance variations. Furthermore, it is suitable both for interior PMSMs and surface PMSMs. Regardless of its promising characteristics, there are a number of issues that can reduce its accuracy, including machine assembling tolerances and rotor lamination grain orientation, which might need to be considered. This paper analyzes the sensitivity of temperature estimation using pulsating d -axis high-frequency current injection to rotor eccentricities, permanent magnet assembling tolerances, and lamination grain orientation. Preliminary results indicate that these issues can have a nonnegligible impact on the accuracy of the method.

Permanent-magnet (PM) magnetization state estimation is important both for torque control and monitoring in conventional permanent-magnet synchronous machines (PMSMs). Furthermore, this can be critical for variable flux machines (VFMs). Use of high-frequency signal injection methods for PM magnetization state estimation in NdFeB magnets has already been proposed. These methods make use of the variation of the PM high-frequency resistance with the PM magnetization state due to the magnetoresistive effect. This paper addresses the generalization of magnetization state estimation using high-frequency signal injection to other types of magnets like SmCo and ferrite, as well as to other magnet structures, e.g., isolated and nonisolated segmented magnets. Use of the magnetoresitive effect for the detection of irreversible/reversible PM demagnetization will also be shown to be viable.

Permanent magnet (PM) temperature measurement/estimation in permanent-magnet synchronous machines (PMSMs) is convenient for both torque control and monitoring purposes. Most of the existing methods provide an averaged/punctual temperature. However, PM temperature distribution is not uniform, which can have unexpected adverse effects such as reduced accuracy estimating the torque produced by the machine or local demagnetization e.g., due to hot spots. PM temperature distribution can be measured or estimated. Inner and outer PM surfaces are not visible once the machine is assembled. Therefore, measurement of the PM temperature distribution cannot be realized by visual means. An array of temperature sensors attached to the PM can be used in this case. While this is technically feasible, it is unaffordable in commercial products due to cost and robustness issues. Alternatively, the PM temperature distribution can be estimated. This paper analyzes the use of back electromotive force (BEMF) harmonics to estimate the differential temperature in the PM of a PMSM, i.e., the temperature difference between the hottest and coldest parts of the magnets. Combined with average PM temperature estimation methods, this information can be used to estimate the PM temperature spatial distribution. The proposed method has been developed on an empirical basis.

This paper analyzes the use of pulsating high-frequency signal injection for islanding detection purposes. Active islanding detection using high-frequency signal injection is an appealing option due to its reduced nondetection zone, reduced cost, and ease of implementation. The use of a rotating high-frequency signal has been reported and analyzed. However, this method can only be applied to three-phase systems. In this paper, the use of a pulsating high-frequency signal injection is proposed. While it uses the same principles as rotating signal injection, it can be applied to both three-phase and single-phase systems, Ministry of Science and Innovation [MICINN-10-CSD2009-00046]; Personnel Research Training Program - Regional Ministry of Education and Science of the Principality of Asturias [BP11-107]

Variable flux permanent magnet synchronous machine (PMSMs) (VF-PMSMs) are gaining importance, especially in automotive applications. The use of such machines will require the development of methods to estimate the permanent-magnet (PM) magnetization state. PM magnetization state estimation methods reported in literature include back-electromagnetic (EMF) compatibility and high-frequency signal injection. One advantage of high-frequency signal injection methods is that they can operate over the whole speed range, with practically no interference with the regular operation of the machine. Magnetization estimation using high-frequency injection relies on the variation of the d- axis high-frequency inductance with the saturation produced by the magnets. Alternatively, the changes of the magnet electrical resistance due to the magnetoresistive effect can also be used for this purpose. These methods have been tested with different materials, including NdFeB, AlNiCo, and SmCo magnets. NdFeB magnets are usually protected from oxidation using NiCuNi coating. However, NiCuNi also shows magnetoresistance effect and can therefore affect the performance of the method when used with magnets having NiCuNi coating. This paper studies the effects of PM coating on the performance of methods, which estimate the magnetization state of the magnet using the magnetoresistance effect by means of high-frequency signal injection. The analysis will include NdFeB, AlNiCo, and SmCo magnets.

Workshop on Control and Modeling for Power Electronics, COMPEL (9th, 2028, Padua), This work was supported by the Spanish Government under projects DPI2014-56358-JIN. It was also supported by the European Commission under FP7 Large Project, grant UE-14-SPEED-604057.

This paper discusses the energization and start-up process of modular three-stage solid-state transformers (SSTs) based on a cascaded H-bridge topology. A key element of the SST is the auxiliary power supply (APS) feeding the auxiliary circuitry. Redundancy and simpler implementation can be achieved by powering the APSs from the dc-link capacitors of cells, but at the expense of a more challenging start-up process of the SST. The first step of this process must be performed before APSs are operative. This can result in undesired transient events (inrush currents, voltage sags in the capacitors voltages, etc.), which could prevent a proper start-up of the SST or even jeopardize the power devices or other elements. A simple strategy for the energization and start-up of SSTs using this APS structure is proposed. The method is valid for both grid-forming and grid-feeding operations of the SST, regardless of the energization port and without the addition of any extra circuitry. Simulation and experimental results are provided.

Permanent-magnet (PM) magnetization state estimation in PM synchronous machines (PMSMs) is of great importance for torque control and monitoring purposes. The magnetization state of a PMSM can change due to several reasons, injection of stator current ( $d$ -axis or $q$ -axis) and variation of magnet temperature being the primary reasons. PM magnetization state estimation is not easy once the machine is assembled. Methods based on the back electromotive force can be used, but they require that the machine is rotating. This paper analyzes the use of high-frequency signal injection for PM magnetization state estimation in PMSMs. The magnetization state of the PMs in PMSMs affects the stator-reflected PM high-frequency resistance. The stator-reflected PM high-frequency resistance can be estimated by injecting a high-frequency voltage/current using the inverter. The high-frequency signal is superposed on the fundamental excitation used to produce torque, meaning that the method can operate at any speed, including zero speed, and without interfering with the normal operation of the machine.

The injection of a high-frequency signal in the stator via inverter has been shown to be a viable option to estimate the magnet temperature in permanent-magnet synchronous machines (PMSMs). The variation of the magnet resistance with temperature is reflected in the stator high-frequency resistance, which can be measured from the resulting current when a high-frequency voltage is injected. However, this method is sensitive to $d$ - and $q$ -axis inductance ( $L_{d}$ and $L_{q} $ ) variations, as well as to the machine speed. In addition, it is only suitable for surface PMSMs (SPMSMs) and inadequate for interior PMSMs (IPMSMs) . In this paper, the use of a pulsating high-frequency current injection in the $d$ -axis of the machine for temperature estimation purposes is proposed. The proposed method will be shown to be insensitive to the speed, $L_{q}$ , and $L_{d} $ variations. Furthermore, it can be used with both SPMSMs and IPMSMs.

Current measurement and pulse-width modulation (PWM) are critical components of modern current controlled voltage source inverters (VSI). Current measurement enables accurate and robust closed-loop control of current, while PWM enables the efficient, high bandwidth regulation of the voltage. These two components do not operate independently, however. The synthesized voltage from PWM, a sequential series of discrete voltage levels, from a predefined finite set of possible levels, is only equivalent to the desired voltage on an average basis. As a result, the response of the electrical current to this synthesized voltage deviates from the response to the desired voltage. This deviation from the desired response, when combined with the standard synchronous sampling method, results in current measurement errors. This paper will present an analytical derivation of this current response deviation. This can be used to model the current measurement errors. A better understanding of this phenomenon will result in better design and control of power converter systems. The analytical work will be evaluated and verified using simulations and experimental measurements.

Permanent magnet synchronous machines (PMSMs) operation above based speed is typically achieved by injecting negative d-axis current to produce flux weakening. However, this increases permanent magnet (PM) losses, reducing therefore the machine efficiency. Previous studies have reported that PM loss increase in flux weakening operation is due to the increase of the airgap flux harmonic content. These studies however did not consider PM properties variation with the MS due to magnetoresistive effect, which can also contribute to PM loss increase. This paper analyzes the variation of PM losses with the MS and flux weakening current. The impact of MS on the thermal behavior of the machine is also studied1.

Rotor eccentricity in permanent magnet synchronous motors (PMSMs) increases unbalanced magnetic pull and motor vibration resulting in accelerated aging of motor components. If eccentricity remains undetected, it can increase in severity, and increase the risk of stator-rotor contact, which causes forced outage of the motor and driven process. Detection of eccentricity currently relies on off-line testing and on-line vibration/current spectrum analysis, which are inconvenient or cannot provide reliable detection as they are influenced by other mechanical non-idealities in the motor or load. In this paper, the feasibility of using the signals from analog Hall-effect field sensors for detecting eccentricity is investigated. It is shown that Hall sensors already present in machines for motion control can be used for measuring the variation in the flux inside the motor due to eccentricity. 3-dimensional (3D) finite element analysis (FEA) and experimental results performed on an interior PMSM (IPMSM) show that the proposed method can provide sensitive and reliable detection of dynamic and mixed eccentricity with minimal hardware modifications.

Torque density of electric motors can be increased if the rotor is placed in the exterior. However, this makes heat extraction more difficult. Study of the thermal behavior at the design stage of the machine can be addressed using FEM and CFD methods. However, they are computationally expensive, and consequently not suitable for the evaluation of multiple designs. A short-time consuming lumped parameter thermal model for external rotor permanent magnet synchronous machines is presented in this work. It allows a fast evaluation of multiple machine design options, being suitable for its use in analytical machine design tools. In addition, comparative analysis of the thermal performance of internal and external rotor designs is provided. The developed model is validated by commercial software achieving good results.

Modular multi-level converters (MMC) are considered a promising topology for HV applications Uke HVDC transmission systems. Such systems must be designed to provide high reliability against different types of faults. MMCs using Half-Bridge (HF) cells are well known for lacking the capability to block short circuits in the DC link. On the contrary, Full-Bridge (FB) cells can block the short circuit current in the DC link, they also enable the implementation of Fault Ride-through (FRT) functionalities, as the cells can generate bipolar voltages in this case. This paper presents a DC short circuit fault ride-through strategy for HVDC transmission system using an MMC topology with FB cells (FBMMC). Two modes of operation are possible during the fault. In the first one, the MMC can support the AC grid with reactive power, but will not transfer active power. In the second mode, the MMC can transfer active power during the fault.

This paper addresses the development of electro-thermal models able to accurately predict the thermal behaviour of power modules in AC drives. The model would be implemented in the digital signal processor used to control the drive, operating therefore on line. FEM will be used for the design and calibration of the models. The performance of the proposed methods will be validated experimentally.

European Conference on Power Electronics and Applications (EPE'17 ECCE Europe) (19th, 2007, Warsaw, Poland), This work was supported by the European Commission FP7 Large Project under grant UE-14-SPEED- 604057 and by the Spanish Government under projects DPI2014-56358-JIN and the grant FPI BES- 2014-070785.

This paper analyzes the dynamic performance of FOC induction machines operating under current and voltage constraints. FOC rely on the proper performance of the inner current control loops. The capability of the machine to increase its torque can be seriously compromised due to the inverter voltage and current limits, which affect to the current regulator transient response. This can be especially critical in machines controlled to operate with reduced rotor flux to increase their efficiency.

The purpose of this paper is to test the algorithm of PWM with inverter non-idealities compensation and switching losses minimization on a model with various power directions and power factor. The PWM strategy implements simultaneous switching losses minimization and dead-time and voltage drop compensation. The operation of the algorithm is considered. Collisions between switching losses minimization and non-idealities compensation are detected.

Solid State Transformers (SSTs) are envisioned as an alternative to conventional Line Frequency Transformers (LFT) able to meed with the requirements of future power systems. Intensive studies covering SSTs' applications, topologies, functionalities and the performance are in progress. Other relevant aspects to consider are the startup (energization) of the transformer as well as its behavior and potential control strategies in the event of anomalies, e.g. imbalances in the load or even short-circuits. This papers addresses all these issues for CHB-based SSTs.

This paper proposes a new predictive torque control algorithm for synchronous reluctance motor drives with the ability to track online the maximum torque per ampere trajectory. An additional term is included in the cost function of the predictive control algorithm which uses an adaptive weighting factor to improve the dynamic behavior of the drive system. As the derivative of torque with respect to the current angle depends on the values of the apparent and incremental inductances, the apparent inductances are estimated online based on the values of the flux linkage and current components while the incremental inductances are estimated using a recursive least squares (RLS) algorithm. Experimental results validate the proposed control algorithm and demonstrate a remarkable performance both in steady-state and during transients, as well as a reduction of the current ripple and audible noise. info:eu-repo/semantics/publishedVersion

This paper proposes the combined use of DPWM strategy to reduce switching losses with inverter non-idealities compensation algorithm. While both methods have been widely studied separately, the combined use poses some restrictions which should be considered. In this paper the priority is given to precise voltage realization and the switching losses optimization is implemented if possible. The proposed method has been implemented and tested in a precision drive. Experimental results are provided.

This paper discusses the use of high frequency (HF) signal injection methods for temperature and magnetization state monitoring of permanent magnet synchronous machines (PMSMs) drives. Temperature and magnetization state are estimated from the HF impedance seen from the stator terminals. A HF signal is injected for this purpose via inverter. These methods does not interfere with the normal operation of the drive, and can be implemented therefore in real time.

This paper proposes a method for islanding detection in microgrids with multiple parallel-connected inverters using high-frequency signal injection. In the proposed method, a master inverter injects the high-frequency signal which is used by the rest of inverters for islanding detection, with two distinguishing features: 1) The slave inverters work in a high-frequency current cancellation mode, what prevents interference and 2) in case of master failure or significant changes in the grid, the remaining inverters will dynamically reassign roles, the new master inverter being self-selected, based on a deterministic performance criteria and without the need of communications, This work was supported in part by the Research, Technological Development and Innovation Programs of the Ministry of Science and Innovation under grant MICINN-10-CSD2009-00046 and of the Spanish Ministry of Science and Innovation-ERDF under grants MICINN-10- ENE2010-14941 and MICINN-10-IPT-370000-2010-15

This paper deals with the thermal stabilization of the power devices which was implemented in the compressor drive of the subway trains. This drive operates in start/stop mode, thus the thermocycling problems in the IGBT modules can occur. The algorithm of thermal stabilization is given and tests were conducted with the real power converter. The thermal images and temperature curves show the efficiency of the thermal stabilization algorithm in comparison to an ordinary solution.

In this paper an adaptive rotor flux observer is developed. This observer performs a real-time correction of the mutual inductance and rotor resistance of the motor using data from the DC-link voltage sensor, the inverter state and the phase current and position sensors. The observer compares the behavior of two independent observers (sensorless and sensored observers) in order to correct the parameters of the sensored observer. The adaptation algorithm corrects the mutual inductance, which can vary due to change of the magnetization current, and the rotor resistance, which can change due to variation of the rotor temperature. Computer simulation results are presented to validate the proposed method.

This article analyzes the effect of the modulation strategy on the thermal behavior of the power devices in AC drives. Excessive temperatures in the Silicon can be prevented by adequate selection of the switching frequency and modulation method. Continuous and discontinuous modulation methods will be considered. The impact of the modulation strategy and switching frequency on the total harmonic distortion (THD) of the currents will also be considered, as increased values of the THD will have adverse effects on the machine, as increased losses, with the subsequent temperature increase. The relationship between the semiconductor temperature (which is not normally measured), and the heat sink temperature (which can be easily measured) will also be analyzed for the different modes of operation. Accurate estimation of the semiconductor temperature will enable advanced, more reliable modes or operation of the power converter.

This article analyzes the effect of the modulation strategy on the thermal behavior of the power devices in AC drives. Continuous and discontinuous modulation methods will be considered. The effects of the switching frequency and of the fundamental frequency on the temperature of the power devices will also be studied.

Nonlinear loads (NLLs) in three-phase systems are known to produce current harmonics at −5, 7, −11, 13,… times the fundamental frequency. However, harmonics of the same frequencies are induced in microgrid voltage thereby reducing the power quality. Dedicated equipment, such as active power filters, can be used to compensate the microgrid harmonics. Alternatively, each distributed generation (DG) unit present in the microgrid can also be used for the same. The use of the virtual admittance concept combined with a proportional integral with resonant controller (PI-RES) control structure has previously been proposed as a harmonic compensation sharing strategy when multiple DGs operate in parallel. The drawback of this methodology is that a large number of RES controllers might be required to compensate for all the harmonic components induced by NLLs, increasing the tuning complexity as well as the execution time. This paper proposes the combined use of virtual admittance control loop and repetitive controller (RC) for harmonic compensation. The main advantage of the proposed method is that only one RC is required to compensate for all the harmonic components, significantly reducing the computational burden and the design complexity. The dynamic performance of the whole system is tested under variable NLL.

The Modular Multilevel Converter (MMC) is a promising technology for medium-high voltage DC/AC converters, being adequate for HVDC transmission systems. Among the appealing characteristics of the MMC are their modularity, and consequently their scalability, as well as the fact that there is no bulk storage element. One key aspect for the operation of the MMC is the response in the event of a short circuit in the DC link. Conventional MMC cells consist of a half-bridge and a capacitor, and have no capability to block the short circuit in the DC side, meaning that expensive and bulky circuit breakers might be needed in this case. Several fault tolerant cell designs have been proposed. However, these desings always bring an increase in the number of power devices and losses. Conventional MMC design can be enhanced to provide added functionalities, e.g. multiport power converters and solid state transformers (SST). A mean to achieve this is by providing the cells the capability to transfer power. This enhancements will imply an increase in the number of power devices and passives, as well as further complexity of the control. However, the resulting cells structures can offer new opportunities regarding fault tolerance. This paper revises the fault tolerance capability of MMCs, and analyzes the behavior of MMC-based multiport power converters in the event of faults. A new cell structure will be proposed capable of blocking the DC short circuit current, therefore protecting the power converter with reduced extra elements.1

Thermal monitoring is a common feature in most of electric machine drives since thermal overloading is one of the most common causes of motor failures. Contact-type sensors are normally used to measure the stator temperature in electric machines. However, use of this type of sensor is not advisable in the rotor as it requires cabling to a rotating part or the use of a wireless transmission system. Consequently, measurement of the rotor temperature is not easy in practice and is not normally implemented in standard machines. An alternative to rotor temperature measurement is rotor temperature estimation. Up to date, only thermal models have been used for rotor temperature estimation in doubly-fed induction machines (DFIMs). This paper proposes rotor temperature estimation in DFIMs using high-frequency signal injection. The proposed method estimates the rotor temperature from the rotor high frequency resistance, which us a function of the rotor windings temperature. The method does not interfere with the normal operation of the drive and can be implemented in existing DFIM drives without requiring additional hardware.1

This paper analyzes a cost-effective modification of the rectifier plus boost converter power topology which is commonly found in small wind turbine systems. The paper proposes removing the input capacitor after the rectifier and the boost converter inductance. This will be done without adding any LC filter between the permanent magnet generator and the passive rectifier. This paper discusses both the power topology and the control strategy as well as a performance comparison with the usual topology. Simulation and experimental results are presented to demonstrate the technical viability of this proposal.

The performance and life expectance of Permanent Magnet Synchronous Machines (PMSMs) is strongly affected by the PM temperature. An increase of the PM temperature reduces the PM strength and consequently the machine torque production capability. Also there is a risk of irreversible demagnetization if PM temperature is too high. Directly measurement of the PM temperature is not easy, PM temperature estimation methods have therefor been widely investigated during the last decade. PM temperature estimation methods can be divided into thermal models, BEMF methods and methods based on the injection of some type of test signal in the stator terminals of the machine. Thermal models require previous knowledge of the machine geometry, materials and cooling system. BEMF and high frequency signal injection methods do not require previous knowledge of the machine geometry or cooling system. However, BEMF methods require that the machine is rotating. High frequency signal injection methods can be used at any speed, including standstill. However, there is a number of issues that can affect to the accuracy of these methods, including stator and rotor designs, machine assembling tolerances and rotor lamination grain orientation. This paper analyzes the sensitivity of high frequency signal based temperature estimation methods to the stator design.1

Conventional line-frequency transformers are a key element in the current power transmission system. Although they are a relatively cheap and well established technology, they are not able to provide new functionalities demanded by the power system operator. Solid State Transformers (SSTs), also called Power Electronic Transformers (PETs) are power electronics-based arrangements able to provide, in addition to the basic functions of a conventional transformer, new functionalities like harmonics, reactive power and imbalance compensation, and power flow control. This paper addresses a comparative analysis of modular PET topologies, including the popular CHB-based approach and MMC-based topologies. Criteria for the evaluation will include aspects like the number of cells required, ratings of the power devices, number and type of the PET ports and design requirements for passive elements.

Islanding detection has been the focus of significant research efforts during the last years. Islanding detection methods can be remote or local, the last being further divided into active and passive methods. Active methods have a reduced non-detection zone (NDZ) and ease of implementation. However they require the injection of some disturbing signal into the grid to measure its response, therefore having a negative impact on the power quality. High frequency signal injection islanding detection methods are a type of active methods that can be used both in single phase and three phase systems. While these methods have been demonstrated to be highly effective in different scenarios, including Single-inverter, Multi-inverter and Microgrid, the analysis reported so far only considered scenarios with linear loads. The proposed paper analyzes the use of high frequency signal injection based islanding detection methods in scenarios including non-linear-loads (NLLs) and active power filters (APFs). The presence of such elements can potentially interfere with the test signals injected for islanding detection; their impact on the method will be studied.1