Please use this identifier to cite or link to this item: http://localhost:8081/xmlui/handle/123456789/15161
Full metadata record
DC FieldValueLanguage
dc.contributor.authorA, Narendrababu-
dc.date.accessioned2021-11-23T06:10:29Z-
dc.date.available2021-11-23T06:10:29Z-
dc.date.issued2018-12-
dc.identifier.urihttp://localhost:8081/xmlui/handle/123456789/15161-
dc.guideAgarwal, Pramod-
dc.description.abstractIn the recent years, the high penetration of various renewable and other non-conventional distributed generators with greater power capability has been an increasing interest. Sources like photovoltaic (PV) generators, fuel cells, wind-energy systems are widely being integrated into power system at distribution level. The main advantages of renewable sources are that they are inexhaustible and environmentally friendly in nature. However, due to uncertainty and uncontrollability of these sources, energy storage systems, power conversion and conditioning techniques are more challenging than ever for grid-connected systems or stand-alone systems. Power electronics, being the technology of efficient electric power conversion is an essential part for integration of various distributed generation systems (renewable energy and non-renewable energy systems) in power system. Power converters such as DC-DC-AC and AC-DC-AC are typically required for integration of low voltage distributed generation sources into grid. Theoretically, a DC-DC boost converter followed by a conventional two-level voltage source inverters (VSI) can increase the voltage to desired level. However, the system imposes several limitations due to various factors and existing semiconductor technology. Since the RMS ac voltage output at the DC-AC input stage is always less than the DC link voltage, several PV arrays are connected in series to meet the required voltage level. A conventional two-level VSI does not allow the independent control of individual PV array in such cases. Multilevel VSIs play the key role in aforementioned scenario by allowing direct connection of several low voltage PV arrays for their independent control and maximum utilization. Over the past few decades, multilevel power conversion technology has generated widespread interest both in the research community and in the industry, as they are becoming a viable technology for many applications. The multilevel topologies like Diode Clamped inverter (DCI), Cascaded H-bridge converter (CHI) and Flying Capacitor inverter (FCI) are considered as conventional multilevel VSIs. They are built with objective of using low voltage rating devices in medium and high voltage applications. However, these conventional multilevel VSI require more switching devices and associated components like driver circuits, power supplies, heatsinks, etc. Therefore, the conventional multilevel VSIs suffers from demerits like complexity, high cost and low reliability compared to two-level VSI counter parts. Fortunately, with substantial improvement of conventional silicon devices and their packaging technologies, new wide band-gap materials reaching higher junction temperature and voltage blocking levels, are being evolved. As a result, voltage rating of the switching devices is no longer a core concern for opting multilevel VSI in various applications. Conversely, ii achieving the functionality of multilevel VSIs (like minimum distortion) with minimum number of switching devices and associated components has become the key factor. This new approach has opened the door to researchers for developing new multilevel topologies aiming different objectives. The research work starts with the literature survey of the multilevel inverter topologies with reduced device count. These topologies have their relative merits and demerits from the stand point of application requirement. No specific topology can be considered as suitable in all respects in most of the applications. Most of the newly proposed topologies in the literature are basically single-phase structures with “polarity-generation unit” and “level-generation unit”. Some of the multilevel topologies need induvial “polarity-generation unit” and “level-generation unit”, while some other can incorporate single “level-generation unit” for three-phase applications. Such inverters with separate “level-generation unit” must incorporate a greater number of DC-sources as the number of phases increases since they cannot be operated with single DC-link. On the other hand, topologies with common “level-generation unit”, redundancy for generating phase voltage levels is limited and it further leads to unequal load sharing among the DC sources. In present work, a detailed investigation is carried out on multipoint clamped inverter (MPCI) topologies, which are realized using single DC link for supplying more than one phase. The DC link of MPCI is formed by several low voltage DC sources like PV/Batteries or a front-end converter system. The output voltage levels of each phase are generated passively through number of clamping points in the common DC link of MPCI topology and hence, it is functionally equivalent to a single‐pole multiple‐throw switch (n ≥ 3 positions) (e.g., the diode-clamped topology). MPCI structure is also advantageous due to the requirement of least number of DC sources for multi-phase operation compared to other topologies. Furthermore, it can generate balanced phase voltages even under DC input source voltage variations, which are likely when using PV/Batteries. This thesis investigates the three-phase MPCI topologies and modulation strategies for the possible applications including renewable energy grid interface, energy storage system, electric drives and energy control centre for smart grid and micro-grids. The key challenges of the power conversion system are identified as power circuit complexity, modulation algorithms and power quality for MPCI topologies. The modulation techniques for MPCIs are relatively complex compared to the two-level counter parts. The control techniques are generally classified as Carrier based modulation and space vector modulation (SVM). Carrier based modulation methods are very simple to implement. On the other hand, SVM provides the complete control over the modulation of the VSI. In this thesis, SVM is considered for detailed analysis. In SVM, there are number of ways to select the set of voltage vectors for synthesizing the reference vector. The selected voltage iii vectors along with the switching sequence play key role for achieving various functionalities of the VSI like: Power loss reduction, power quality, capacitor voltage balancing etc. In the present work, the SVM strategies are classified as nearest three vector modulation (NTV) and non-nearest three voltage vector modulation (non-NTV). NTV modulations depending on the type voltage vectors selected. NTV strategies are very common as they produce better harmonic performance. However, depending on the MPCI topology, NTV strategies have several limitations. In such cases, non-NTV methods are essential even if they relatively increased harmonic distortion in the output voltage and current waveforms. Three-level (3L) neutral point clamped inverter (NPCI) is the simplest example of MPCI. These are proven to be excellent tradeoff solution between performance and cost in high-voltage and high-power systems. The main advantages of 3L NPCI are reduced voltage ratings for the switches, good harmonic spectrum (making possible the use of smaller and less expensive filters), and good dynamic response. Even for low voltage applications, the NPCI and equivalent topologies like T-type NPCI are widely used as an alternative to two-level VSIs. In the present work, first 3L NPCI is presented thoroughly. NTV and non-NTV modulation strategies are implemented on 3L NPCI for elimination of low frequency neutral point (NP) voltage oscillations. A simplified implementation algorithm is developed for NTV and non-NTV modulation strategies. Therefore, coordinate transformation, trigonometric expressions and solutions of volt-second balance equations are not required in the proposed algorithms. Moreover, the unique feature the proposed modulation algorithm is that, they do not require independent voltage-second balance equations for NTV and non-NTV modulation strategies. Therefore, these algorithms can be easily modified for implementing the hybrid modulation methods without increasing the computational burden. Further analysis is focused on reducing the complexity of MPCI topologies by minimizing power semiconductor device (PSD) count. In this context, first, a reduce device count 3L inverter termed as hybrid 2/3 level (2/3L) NPCI is presented based on the concept of non-NTV modulation. It can be formed by addition of two half bridge modules to a two-level inverter thereby uses only ten active switches. It can save two active switches and six clamping diodes compared to the conventional NPCI. However, as a result of reduction in switching combinations, the medium voltage vectors are absent in the SVD of hybrid 2/3 level inverter. The voltage vector selection is also no longer similar to conventional 3L NPCI. In order to address this issue, two types of voltage vector selection methods for hybrid 2/3L NPCI are investigated to approximate the reference vector very closely. Effect of different switching state selections is studied in detail. A new virtual vector modulation to use both the small vectors for entire modulation index range is also proposed. Therefore, the modulation allows the interconnection of Z-source network to the hybrid 2/3L NPCI. Two new switching sequences for iv higher modulation range are analyzed to retain the same number of device commutations even with the insertion of shoot-through states. It was observed that, in case of MPCI, if an attempt is made to reduce the number of active switches, it affects the number of available switching states (or voltage vectors) for synthesizing the reference voltage vector. Therefore, such devices are considered as reduced switching state (RSS) MPCI topologies. The position of available voltage vectors in SVD of a given RSS MPCI depends on i) total number of switching devices and ii) structure of the RSS MPCI topology. This effect is further explored by studying four and five level RSS MPCI topologies in detailed with different possible structures. Modulation techniques for such topologies are also investigated for synthesizing the reference vector using the available voltage vectors for achieving minimum distortion. Computer simulation studies have been carried out in MATLAB/Simulink to verify the performance of various MPCI topologies and modulation strategies under different load conditions. The simulation study has been carried out for both steady-state and transient conditions. To validate the simulation studies of different topologies, downscaled hardware prototypes have been designed, developed and tested. The Real-Time Interface (RTI) of MATLAB and RT-Lab controller (OP 5600) are used to generate gate pulses for switching devices of the developed prototype. Different hardware components as required for the operation of the experimental set-up such as pulse amplification, isolation circuit, dead-band circuit, voltage and current sensor circuits have been developed and interfaced with the RT-Lab controller. As the experimentation is carried out at reduced voltage, for validating the experimental results, downscale simulation results are also performed and compared with experimental results. The experimental results have been found to be in good agreement with the simulation results. Although all the proposed schemes are experimentally tested and validated on downscaled laboratory prototypes, but the proposed topologies, SVM technique are general in nature and can be easily applied to high power applications.en_US
dc.description.sponsorshipIndian Institute of Technology Roorkeeen_US
dc.language.isoenen_US
dc.publisherI.I.T Roorkeeen_US
dc.subjectGreater Power Capabilityen_US
dc.subjectWind-Energy Systemsen_US
dc.subjectPower Convertersen_US
dc.subjectFortunatelyen_US
dc.titleTHREE PHASE MULTIPOINT CLAMPED INVERTERS AND MODULATION STRATEGIESen_US
dc.typeThesisen_US
dc.accession.numberG28713en_US
Appears in Collections:DOCTORAL THESES (Electrical Engg)

Files in This Item:
File Description SizeFormat 
G28713.pdf18.19 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.