Browsing by Author "Xu, Yue"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
- Deep Recurrent Q Networks for Dynamic Spectrum Access in Dynamic Heterogeneous Envirnments with Partial ObservationsXu, Yue (Virginia Tech, 2022-09-23)Dynamic Spectrum Access (DSA) has strong potential to address the need for improved spectrum efficiency. Unfortunately, traditional DSA approaches such as simple "sense-and-avoid" fail to provide sufficient performance in many scenarios. Thus, the combination of sensing with deep reinforcement learning (DRL) has been shown to be a promising alternative to previously proposed simplistic approaches. DRL does not require the explicit estimation of transition probability matrices and prohibitively large matrix computations as compared to traditional reinforcement learning methods. Further, since many learning approaches cannot solve the resulting online Partially-Observable Markov Decision Process (POMDP), Deep Recurrent Q-Networks (DRQN) have been proposed to determine the optimal channel access policy via online learning. The fundamental goal of this dissertation is to develop DRL-based solutions to address this POMDP-DSA problem. We mainly consider three aspects in this work: (1) optimal transmission strategies, (2) combined intelligent sensing and transmission strategies, and (c) learning efficiency or online convergence speed. Four key challenges in this problem are (1) the proposed DRQN-based node does not know the other nodes' behavior patterns a priori and must to predict the future channel state based on previous observations; (2) the impact to primary user throughput during learning and even after learning must be limited; (3) resources can be wasted the sensing/observation; and (4) convergence speed must be improved without impacting performance performance. We demonstrate in this dissertation, that the proposed DRQN can learn: (1) the optimal transmission strategy in a variety of environments under partial observations; (2) a sensing strategy that provides near-optimal throughput in different environments while dramatically reducing the needed sensing resources; (3) robustness to imperfect observations; (4) a sufficiently flexible approach that can accommodate dynamic environments, multi-channel transmission and the presence of multiple agents; (5) in an accelerated fashion utilizing one of three different approaches.
- Design of a 10 kV SiC MOSFET-based high-density, high-efficiency, modular medium-voltage power converterMocevic, Slavko; Yu, Jianghui; Fan, Boran; Sun, Keyao; Xu, Yue; Stewart, Joshua; Rong, Yu; Song, He; Mitrovic, Vladimir; Yan, Ning; Wang, Jun; Cvetkovic, Igor; Burgos, Rolando; Boroyevich, Dushan; DiMarino, Christina; Dong, Dong; Motwani, Jayesh Kumar; Zhang, Richard (IEEE, 2022-03)Simultaneously imposed challenges of high-voltage insulation, high dv/dt, high-switching frequency, fast protection, and thermal management associated with the adoption of 10 kV SiC MOSFET, often pose nearly insurmountable barriers to potential users, undoubtedly hindering their penetration in medium-voltage (MV) power conversion. Key novel technologies such as enhanced gatedriver, auxiliary power supply network, PCB planar dc-bus, and high-density inductor are presented, enabling the SiC-based designs in modular MV converters, overcoming aforementioned challenges. However, purely substituting SiC design instead of Sibased ones in modular MV converters, would expectedly yield only limited gains. Therefore, to further elevate SiC-based designs, novel high-bandwidth control strategies such as switching-cycle control (SCC) and integrated capacitor-blocked transistor (ICBT), as well as high-performance/high-bandwidth communication network are developed. All these technologies combined, overcome barriers posed by state-of-the-art Si designs and unlock system level benefits such as very high power density, high-efficiency, fast dynamic response, unrestricted line frequency operation, and improved power quality, all demonstrated throughout this paper.
- Development of Nation Wide Cost-Benefit Analysis Framework for Aviation Decision Making Using Transportation Systems Analysis ModelXu, Yue (Virginia Tech, 2008-03-28)The aim of this study is to establish a nation-wide cost-benefit framework for aviation projection appraisal. This framework is built upon Transportation System Analysis Model developed at Virginia Tech Air Transportation System Model (TSAM). Both supply and demand characteristics and their inter-dependence are investigated. It attempts to solve the absence of supply constraints in aviation demand forecast in the literature. In addition, external costs in term of noise and emission are also considered. A national environmental impact analysis introduced by new generation small aircraft system is conducted. Two case studies are discussed to illustrate the framework. The first one is based on the GPS Wide Area Augmentation System (WAAS) Lower Landing Minima capability. It represents a nation-wide cost-benefit analysis with examination of both supply and demand. System-wide benefit of accessibility improvement and infrastructure cost are scrutinized at the same time. A prioritized set of candidate airports for this technology is provided as a result. The second study focuses on New York area. Benefits brought by DataComm technology are evaluated by multi-iteration simulations. DataComm is projected to reduce entry point intrail and final approach separation. The improvements are modeled at individual airport and New York airspace. Consumer surplus is estimated based on demand and delay relationship using TSAM.
- Insulation Design, Assessment and Monitoring Methods to Eliminate Partial Discharge in SiC-based Medium Voltage ConvertersXu, Yue (Virginia Tech, 2021-07-07)In comparison with Si-based converters, the emerging Medium Voltage (MV) SiC-based converters can achieve higher blocking voltage capability, lower on-resistance, faster switching speed with less switching related losses and run under higher temperature. Thus, theoretically, it can achieve much higher power density, which becomes very promising for future power transmission and distribution. However, in order to achieve the desired high power density, insulation system of the MV SiC-based converter must be compact. Therefore, challenges for the insulation system gradually appeared, as the insulation size becomes smaller and the Electric field (E-field) intensity significantly increases. Under such high E-field intensity, it is necessary and important to eliminate Partial Discharge (PD) for such power converters, since the converter system is vulnerable to PDs. Developing an insulation design, assessment and online monitoring method to help reach a compact and PD free insulation system for MV SiC-based converters is a goal of this work. General insulation design and assessment guidelines based on experimental PD investigation and physics-based model –Experimental PD investigation is completed for internal void discharge, surface discharge, and point discharge representative coupons under square excitations. Based on the data and the existing knowledge about PD mechanisms, widely accepted PD models are selected. Using these physics-based models, simulation results can demonstrate the major features observed in the experiments. With the experimental data and valid PD models, several general insulation design and assessment guidelines are proposed, which could be further applied during converter prototypes development. Partial Discharge elimination methodology and design examples – By using the laminated bus as the design example, internal void evaluation and analysis method is demonstrated. Then, targeting the internal PD-free design with reasonable insulation thickness, several insulation improvement methods are applied and experimentally verified by using representative coupons. After understanding the possible ways for evaluating and eliminating internal voids, a PCB-based planar bus is designed and fabricated, which shows great insulation improvement after experimental verification. In order to eliminate PDs in the air and shrink the insulation distance, three ways for managing E-field distribution in air are demonstrated by three examples. First, by using the interconnections among the power modules, Rogowski-based current-sensing board, and the laminated bus as an example, E-field distribution can be estimated by Finite Element Analysis (FEA) and its management can be achieved by geometrical modifications. Second, for the one-turn inductor, a methodology is demonstrated that builds a coaxial insulation structure with proper termination technology in order to squeeze air out of the insulation system. Finally, E-field shielding technology is applied along the heatsink edges in order to make the E-field distribution uniform and to shrink the insulation distance between the heatsink and the cooling system. After improving the insulation, this work shrinks the converter unit size by around 50% while maintaining its PD-free status under normal operation conditions. Besides the significant increase in power density and weight reduction, the entire converter system has less ringing and better current-sharing performance due to reductions of the parasitic inductance. Partial Discharge online monitoring via acoustic and photon detection methods –Targeting the online monitoring and even localization of surface discharge for power converter applications, two novel types of sensors have been proposed and fabricated. In order to verify the concepts, one example with experimental results has been given for each type of sensor. The experimental data demonstrates that such sensors can be placed inside the converter and online monitoring can be realized for surface or corona discharges by capturing either the acoustic signal or the photons that are generated by discharge events.