Table of contents monday, September 9, 1: 30pm-4: 00pm modular Multi-Level Converters, hvdc, and dc grids I 3

P3105 High Dynamic Sensorless Control for PMSMs Based on Decoupling Adaptive Observer [#581]
Yongle Mao, Jiaqiang Yang, Tao Wang, Dejun Yin and Yangsheng Chen, Zhejiang University, China; Nanjing University of Science and Technology, China

Voltage reference that contains cross-coupling term between the q- and d-axis is commonly utilized as input of sensorless schemes, and this might cause coupling between speed and position estimation and deteriorate system dynamic performance, which is unacceptable in some applications such as electric vehicle and industrial robot. By employing decoupled voltage reference as observer input, this paper proposes a decoupling adaptive observer for permanent magnet synchronous motor (PMSM), with the estimated rotor speed and position regulated by current estimation errors of the q- and d-axis independently and simultaneously. Dynamic performance of the speed and position estimators, which can be treated as two separate Single-Input-Single-Output (SISO) systems, is analyzed based on their transfer functions, and explicit guidelines for observer gains determination are derived. Experimental tests are conducted on a 1.0 kW PMSM testing platform to verify the effectiveness of the proposed sensorless control approach.

P3106 Position Sensorless Control of Switched Reluctance Motor Based on a Numerical Method [#683]
Fei Peng, Jin Ye and Ali Emadi, McMaster University, Canada; San Francisco State University, United States

In this paper, a new position sensorless control method for switched reluctance motor drives is proposed. Rotor position is initially calculated based on the flux linkage-position- phase current characteristics by numerical method. Then, a third-order phase locked loop considering the acceleration variation is designed to undermine the impact of current sampling noise and numerical residual error on the estimated rotor position. Simulation and experimental results show that the proposed position sensorless control method has achieved sufficient accuracy in terms of position and speed estimation.

P3107 Operating-Envelop-Expandable Control Strategy for Switched Flux Hybrid Magnet Memory Machine [#1042]
Hui Yang, Heyun Lin, Z. Q. Zhu, Erxing Zhuang, Shuhua Fang and Yunkai Huang, Southeast University, China; University of Sheffield, United Kingdom

Memory machines (MMs) equipped with hybrid permanent magnets (PMs), i.e., NdFeB and low coercive force (LCF) PMs, combine the merits of acceptable torque capability at low speeds and efficiency improvement at high speeds. Meanwhile, the PM flux linkage can be flexibly adjusted with the aid of current pulses, which is desirable for variable speed applications. To effectively extend the constant power speed range (CPSR) with minimum required flux-weakening (FW) current injected by the inverter, a new control strategy is proposed and implemented on a switched flux hybrid magnet memory machine (SF-HMMM). The configuration and operating principle of the proposed machine are introduced first, followed by the establishment of the mathematical model. Afterwards, the proposed control strategy is described accounting for a whole operating envelop. Different FW regions will be numerically identified by calculating the transition speed. In FW region I, the conventional d-axis current-based and PM demagnetization FW schemes are integrated, while only the negative d-axis current injection is employed in FW region II. The proposed control strategy combines the distinct advantages of memorable flux and negative Id FW controls, leading to an effective extension of CPSR with improved efficiency as well as low requirement of inverter current. The proposed method is verified by experiments on the prototype machine.

Poster Session: Power Semiconductor Devices, Passive Components, Packaging, Integration, and Materials

Tuesday, September 20, 11:00AM-12:30PM, Room: Exhibit Hall, Chair: Giovanna Oriti, Enrico Santi

P3301 Modelling the closely coupled cascode switching process [#1505]
Pablo F. Miaja, Sheng Jiang, Kean-Boon Lee, Peter A. Houston, Ivor Guiney, David J. Wallis, Colin J. Humphreys and Andrew J. Forsyth, University of Manchester, United Kingdom; University of Sheffield, United Kingdom; University of Cambridge, United Kingdom

The cascode combination of transistors is a simple and interesting way of turning a normally-on device into a normally-off one. Cascodes are usually used with wide bandgap materials such as gallium nitride (GaN) and silicon carbide (SiC) in which a wide bandgap normally-on device is connected to a silicon MOSFET. Integration of the different transistors to form a closely-coupled cascode leads to an improvement in the switching characteristics. By undertaking a detailed analysis of the closely-coupled cascode switching behaviour under constant current drive conditions, an insight into the role of the different parasitic capacitances and the current capability (related to the size) of the transistors can be obtained. This is of great importance when designing all-GaN die-integrated cascodes. Also, it provides a basis for comparing cascodes with single devices in terms of the Qg.Ron figure of merit.

P3302 A 700-V Class Reverse-Blocking IGBT for Large Capacity Power Supply Applications [#159]
David Hongfei Lu, Hiromu Takubo, Hiroki Wakimoto, Toru Muramatsu and Haruo Nakazawa, Fuji Electric Co., Ltd., Japan

A 700-V class Reverse-Blocking IGBT (RB-IGBT) is developed for large capacity power supply applications inevitably accompanied with increasing parasitic inductance and high current slew rate relative to their medium capacity counterparts. Compared with 600-V RB-IGBT, the device allows high dynamic surge voltage and fast switching to reduce turn-off loss by 35% under an advanced T-type-NPC three level power module benchmark condition and by 17% from a loss analysis of a realistic three level rectifier stage in a generic power converter application. Furthermore, reduction of reverse-blocking leakage current to one-tenth also enables 25 deg. C higher operation junction temperature. A side effect of the developed device as a first step is a 0.15V increase in on-state voltage, making it gain no loss reduction in a three level inverter stage. Therefore, its reduction should be addressed in further work.

P3303 Efficiency and Electromagnetic Interference Analysis of Wireless Power Transfer for High Voltage Gate Driver Application [#1044]
Jianyu Pan, Feng Qi, Haiwei Cai and Longya Xu, The Ohio State University, United States

Wireless power transfer (WPT) holds great potential to achieve strong galvanic isolation, compact size, and low parasitic capacitance in the power supply of gate driver. In this paper, a wireless power supply is designed and analyzed for high voltage gate driver application. The performance of the power supply in terms of power delivery, parasitic capacitance, and driver output is explored in the functional test.Meanwhile, electromagnetic interference (EMI) is investigated in both frequency and space distribution based on measurement and simulation. The experimental results demonstrate that the prototype successfully delivers power for gate driver board with an 80 mm air gap while the maximum transfer efficiency is kept around 90%. The parasitic capacitance brought by the air gap is only 1.72 pF, which significantly increases common mode impedance. Radiated EMI affects the frequency span around the resonant frequency, and its distribution regularity in space is clearly displayed by field curves in horizontal axis and vertical axis.

P3304 Single Chip Enabled High Frequency Link based Isolated Bias Supply for Silicon Carbide MOSFET Six-Pack Power Module Gate Drives [#1680]
Rui Gao, Li Yang, Wensong Yu and Iqbal Husain, North Carolina State University, United States

Regarded as one of the most successful wide bandgap (WBG) devices, Silicon Carbide (SiC) metal oxide semiconductor field transistors (MOSFETs) are being considered in an increasing number of power electronics applications. One of those applications is the hybrid and electric vehicle (HEV EV) traction inverters where high efficiency and high power density is essential. From the system level perspective, the gate driver circuit design for such device is challenging considering the device's fast switching speed and compact system structure. This paper presents a low profile isolated bias supply design using commercially available components for the SiC MOSFET modules targeting an HEV EV traction inverter application. A single chip MAX 13256 is adopted to form the high frequency link for entire power module gate drive supply. Distributed transformer strategy is highlighted to provide multiple isolated output and compact structure with minimized parasitic capacitance between all the isolation barriers. The featured low profile optimization reduces the parasitic parameters that might deteriorate the system performance for the fast switching WBG devices. Moreover, the open-loop high-frequency link architecture allows easy configuration for customized output voltage level, polarity and higher reliability. A prototype gate driver has been built for 1.2 kV, 50 A SiC six pack MOSFET power module, and experimental results are presented.

P3305 Reliability Assessment of SiC Power MOSFETs From The End Users Perspective [#91]
Vasilios Dimitris Karaventzas, Muhammad Nawaz and Francesco Iannuzzo, Dept. of Energy Technology, Aalborg University, Denmark; ABB Corporate Research, Sweden

The reliability of commercial Silicon Carbide (SiC) Metal-Oxide-Semiconductor Field-Effect Transistors (MOSFETs) is investigated, and comparative assessment is performed under various test environments. The MOSFETs are tested both regarding the electrical properties of the dies and the packaging properties of the devices. The results of each reliability stress are utilized not only for mutual comparison of SiC-based power commercial modules, but also as a tool to understand the underlying physical mechanisms of degradation. Towards this goal, the devices were placed under accelerate stress conditions, such as: high electric field, high temperature and high humidity. Finally, a preliminary judgment is performed on each kind of stress, based on the quality assessment of the semiconductor as well as the packaging material.

P3306 Investigation of Collector Emitter Voltage Characteristics in Thermally Stressed Discrete IGBT Devices [#1296]
Syed Huzaif Ali, Serkan Dusmez and Bilal Akin, University of Texas at Dallas, United States

Discrete package Insulated Gate Bipolar Transistor (IGBT) devices are a popular choice for low-medium power converters. Although IGBT power modules have been extensively studied in literature, there exists a major gap for reliability study particularly for discrete devices. Current failure diagnostic tools are not mature enough for failure diagnosis and prognosis in real-time operation based on system condition monitoring. In order to move power conversion technologies forward reliably, this paper investigates the changes in on-state collector- emitter voltage drop (Vceon) of discrete IGBT devices exposed to thermal cyclic stress. Depending on the dominant aging mechanisms and device structure, Vceon variation trend is different. In this paper, Vceon variations of three different IGBT device types, namely, punch through, non-punch through and field stop, are continuously monitored during accelerated aging tests, and their relations to aging mechanisms are identified.

Poster Session: Emerging Technologies and Applications

Tuesday, September 20, 11:00AM-12:30PM, Room: Exhibit Hall, Chair: Jin Wang, Yaosuo "Sonny" Xue

P3501 Transmission Characteristics Analysis of a Three-Phase Magnetically Coupled Resonant Wireless Power Transfer System [#749]
Jiang Chong, Liu Fuxin, Ruan Xinbo and Chen Xuling, Nanjing Univ. of Aeronautics and Astronautics, China

Multi-phase magnetically coupled resonant (MCR) wireless power transfer (WPT) technology has attracted wide spread attention recently, for it can not only reduce the sensitivity of system to the spatial scale effectively, but also maintain the high transmission efficiency and power in the middle distance. This paper presented an analytical equivalent model for a three-phase MCR WPT system under phase angle control method with different spatial locations between the sending coils and receiving coil. The mutual inductance formulas of the sending coils and receiving coil were derived. The relationship among the output power, transmission efficiency and the angular misalignments were analyzed in detail. Experiments have also been carried out to facilitate quantitative comparison and validate the theoretical analysis.

P3502 Synthesis of Buck Converter Based Current Sources [#866]
Soumya Shubhra Nag and Santanu Mishra, Indian Institute of Technology Kanpur, India

Ripple free DC or AC current source can be implemented by operating a switching converter in current mode control and setting the switching frequency much higher than the operating frequency. However, due to presence of the output filter across the load, it exhibits deviation from the ideal output impedance characteristics of an ideal current source. In this paper, a DC and an AC current source are presented which provides low ripple output current without sacrificing the high output impedance characteristics of an ideal current source. The DC and AC current sources are realized using a buck converter and a voltage source inverter, respectively with a coupled LLLC output filter replacing the traditional LC output filter. The low ripple output current is obtained by utilizing the notch behavior of the coupled LLLC filter. The operating principle of the proposed current source structures are explained in details. The effects of passive component tolerance on the current ripple attenuation and output impedance is also discussed in details. The performances of the proposed current sources are validated using simulations and experiments. The proposed DC current source and sinusoidal AC current source shows very low ripple content at their output current.

P3503 A Model for Coupling Under Coil Misalignment for DD Pads and Circular Pads of WPT Systems [#868]
Guangjie Ke, Qianhong Chen, Ligang Xu, Siu-Chung Wong and Chi.K. Tse, Nanjing University of Aeron. and Astro., China; Nanjing University of Aero. and Astro., China; Hong Kong Polytechnic University, Hong Kong

The coupling coefficient of a transformer is a key parameter affecting the performance of a wireless power transmission system. Based on the magnetic reluctance model under aligned condition, this paper gives an effective method for calculating the null coupling position, as well as a practical and accurate model for finding the coupling coefficient under coil misalignment and varying air gap. The null coupling position is determined by the horizontal offset when the flux of opposite direction that is intersected with the receiving coil is cancelled out. An empirical formula is given to correct the null coupling position. Based on the physics of coupling under misaligned conditions, a practical model for coupling is constructed. Transformers with DD structure and typical circular structure are fabricated for verification. Experimental measurement validates the model.

P3504 Comprehensive Dynamic Modeling of a Solid-state Transformer Based Power Distribution System [#1129]
Md Tanvir Arafat Khan, Alireza Afiat Milani, Aranya Chakrabortty and Iqbal Husain, North Carolina State University, United States

This paper presents a physics based comprehensive dynamic model of a future power distribution system, termed as the FREEDM system, for plug-and-play interface of distributed renewable energy resources and distributed energy storage devices. The system allows for high penetration of renewable generation with energy storage at the distribution level. FREEDM system consists of an energy router, which is the power electronics based solid- state transformer (SST) that interfaces distributed generation, storage and local loads on the low voltage side with the medium voltage node of the distribution grid. In this paper, state-space modeling and dynamic performance of the SST is analyzed along with the renewable generation sources and storage components with the goal of studying the feasible operating points of the FREEDM system. The actual model of the single-SST system amounts to highly complex dynamics with more than hundred state variables. Singular perturbation based model reduction techniques are applied, thereby leading to a 70th order state-space average model suitable for AC and DC energy cell system sizing, stability analysis, and controller design. The analysis with the system model revealed the SST input stage system parameters have the dominant effect on the feasible operation region.

P3505 Capability, Compatibility, and Usability Evaluation of Hardware-in-the-Loop Platforms for DC-DC Converter [#1198]
Shawn Maxwell, S M Rakiul Islam, Md. Kamal Hossain and Sung Yeul Park, University of Connecticut, United States

This paper evaluates the capability, compatibility, and usability of Hardware-in-the-Loop platforms for DC-DC converter. This was accomplished by interfacing the platforms with a physical power stage as well as a controller. The employed platforms are Hi-Rel Power-pole board, Texas Instruments Digital Controller, RTDS, OPAL-RT, dSPACE and Typhoon. Two sets of experimentation were performed: the power stage represented by the Power-pole board, RTDS, OPAL-RT, dSPACE, and Typhoon and the controller replaced by TI DSC, RTDS, OPAL-RT, dSPACE, and Typhoon. Three points of evaluation for a testing platform that are of interest to industrial researchers as well as academia are capability (speed or modeling capacity), compatibility (ease of porting models from other platforms), and usability (ease of use of software and hardware). This paper provides an introductory resource for research and education by providing results of a simple buck converter example.

P3506 A Single Stage AC/DC Converter for Low Voltage Energy Harvesting [#1230]
Liang Yu and Haoyu Wang, ShanghaiTech University, China

In this paper, a novel ac/dc converter is proposed for low voltage, low power rectification applications. The proposed converter manages the energy harvested from micro-scale electromagnetic transducers. It integrates the conventional Boost and Buck-Boost topologies with a shared inductor, a bidirectional switch and two split filtering capacitors. The Boost and the Buck-Boost topologies function in the positive and negative half input cycles, respectively. The inductor is energized by being shorted with the input source through the MOSFET channel without using the diodes. This enables active rectification of low amplitude (below 0.7 V) ac voltages. Theoretical analysis, design considerations and control method are detailed. A 45 mW circuit prototype, which converts a 0.4 V peak, 100 Hz ac voltage to 3.3 V dc is designed and tested to verify the proof of concept.

Tuesday, September 20, 3:00PM-4:30PM

Poster Session: Renewable and Sustainable Energy Applications

Tuesday, September 20, 3:00PM-4:30PM, Room: Exhibit Hall, Chair: Euzeli Santos Jr., Johan Enslin

P3701 Dynamic Battery Operational Cost Modeling for Energy Dispatch [#1549]
Qian Zhao, Aniq Ahsan, Ashwin M. Kambadkone and Meng Hwee Chia, Experimental Power Grid Center, Singapore; Oxford University, United Kingdom; National University of Singapore, Singapore

Battery Energy Storage Systems (BESS) have gained extensive application in both grid and microgrid applications. One major type of BESS are electrochemical batteries such as Lead-Acid and Lithium-Ion batteries which have limited number of lifecycles. The common way of considering their operation cost is using a constant value such as LCOE (levelized cost of energy). However, as shown herein, given the same amount of energy output, the battery lifecycle degradation, and thus the degradation cost, can vary at different operation conditions (voltage, current, power, state of charge (SOC)) by up to 6 times. Herein a model for the dynamic battery operation cost as a function of its dispatch power and SOC is developed. The model also considers the dependency of battery voltage on its current and SOC, which equivalently takes into account the dependency of its conversion efficiency on its power and SOC. Preliminary simulations demonstrate that using the proposed model, instead of the LCOE, for Microgrid operation optimization microgrid operation cost is lower by up to 12%.

P3702 A Low Voltage Ride Through Control Strategy for Energy Storage Systems [#349]
Yeongsu Bak, June-Seok Lee and Kyo-Beum Lee, Ajou University, Korea (South); KRRI, Korea (South)

This paper proposes a low voltage ride through (LVRT) control strategy for energy storage systems (ESSs). The LVRT control strategies for wind turbine systems and photovoltaic systems have been researched until now. Regardless of the energy source, the main aim of the LVRT control strategies for a grid side converter is to inject the reactive power according to the gird code regulations. The main aim of the proposed LVRT control strategy for ESSs is the same as them; however, it additionally considers the case of charging state which cannot be taken into consideration in wind turbine systems and photovoltaic systems having unidirectional power flow. The proposed LVRT control strategy for ESSs determines not only the reactive reference current for injecting the reactive power but also the active reference current to contribute to a point of common coupling (PCC) voltage increase by considering the two operating condition of charging and discharging state. The validity of the analysis is verified by simulation results.

P3703 Experimental Validation of the Solid State Substation with Embedded Energy Storage Concept [#1010]
Christian Klumpner, Mohamed Rashed, Dipankar De, Chintan Patel, Ponggorn Kulsangcharoen and Greg Asher, University of Nottingham, United Kingdom

This paper proposes the concept of integrating the energy storage within a Medium Voltage to Low Voltage solid state substation in order to provide new features compatible with the requirements from future more intelligent grids. The principles for the system development are presented and the role of each subsystem is explained. The experimental evaluation of the 1.9kVrms/25kVA substation with 1.5MJ of supercapacitor storage consists of subsystem tests showing the waveform quality, step transients as well as system tests of the efficiency, ride-through and power peak shaving operation.

P3704 Understanding Dynamic Model Validation of a Wind Turbine Generator and a Wind Power Plant [#151]
Eduard Muljadi, Yingchen Zhang, Vahan Gevorgian and Dmitry Kosterev, National Renewable Energy Laboratory, United States

Regional reliability organizations require power plants to validate the dynamic models that represent them to ensure that power systems studies are performed to the best representation of the components installed. In the process of validating a wind power plant (WPP), one must be cognizant of the parameter settings of the wind turbine generators (WTGs) and the operational settings of the WPP. Validating the dynamic model of a WPP is required to be performed periodically. This is because the control parameters of the WTGs and the other supporting components within a WPP may be modified to comply with new grid codes or upgrades to the WTG controller with new capabilities developed by the turbine manufacturers or requested by the plant owners or operators. The diversity within a WPP affects the way we represent it in a model. Diversity within a WPP may be found in the way the WTGs are controlled, the wind resource, the layout of the WPP (electrical diversity), and the type of WTGs used. Each group of WTGs constitutes a significant portion of the output power of the WPP, and their unique and salient behaviors should be represented individually. The objective of this paper is to illustrate the process of dynamic model validations of WTGs and WPPs, the available data recorded that must be screened before it is used for the dynamic validations, and the assumptions made in the dynamic models of the WTG and WPP that must be understood. Without understanding the correct process, the validations may lead to the wrong representations of the WTG and WPP modeled.