Browsing by Author "Ahmed, Sara Mohamed"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
- Computer Modeling and Simulation of Power Electronics Systems for Stability AnalysisAhmed, Sara Mohamed (Virginia Tech, 2007-12-05)This works focuses on analyzing ac/dc hybrid power systems with large number of power converters that can be used for a variety of applications. A computer model of a sample power system is developed. The system consists of various detailed/switching models that are connected together to study the sample system dynamic behavior and to set conditions for safe operation. The stability analysis of this type of power systems has been approached using time domain simulations. There are three types of stability analysis that are studied: steady-state, small-signal analysis and large signal analysis. The steady-state stability analysis is done by investigating the nominal operation of the power electronics system proposed. The small-signal stability of this system is studied by running different parametric case studies. First, the safe values of the main system parameters are defined from the view of the stability of the complete system. Then, these different critical parameters of the system are mapped together to predict their influence on the system. The large signal stability is examined through the response of the power system to different types of transient changes. There are different load steps applied to the critical parameters of the system at the maximum or minimum stability boundary limit found by the mapping section. The maximum load step after which the system can recover and remain stable is defined. The other type of large signal stability analysis done is the study of faults. There are different faults to be studied; for example, over voltage, under voltage and over current.
- Modeling of Power Electronics Distribution Systems with Low-frequency, Large-signal (LFLS) ModelsAhmed, Sara Mohamed (Virginia Tech, 2011-04-29)This work presents a modeling methodology that uses new types of models called low-frequency, large-signal models in a circuit simulator (Saber) to model a complex hybrid ac/dc power electronics system. The new achievement in this work is being able to model the different components as circuit-based models and to capture some of the large-signal phenomena, for example, real transient behavior of the system such as startup, inrush current and power flow directionality. In addition, models are capable of predicting most low frequency harmonics only seen in real switching detailed models. Therefore the new models system can be used to predict steady state performance, harmonics, stability and transients. This work discusses the modeling issues faced based on the author recent experiences both on component level and system level. In addition, it recommends proper solutions to these issues verified with simulations. This work also presents one of the new models in detail, a voltage source inverter (VSI), and explains how the model can be modified to capture low frequency harmonics that are usually phenomena modeled only with switching models. The process of implementing these different phenomena is discussed and the model is then validated by comparing the results of the proposed low frequency large signal (LFLS) model to a complete detailed switching model. In addition, experimental results are also obtained with a 2 kW voltage source inverter prototype to validate the proposed improved average model (LFLS model). In addition, a complete Verification, Validation, and Uncertainty Quantification (VV&UQ) procedures is applied to a two-level boost rectifier. The goal of this validation process is the improvement of the modeling procedure for power electronics systems, and the full assessment of the boost rectifier model predictive capabilities. Finally, the performance of the new models system is compared with the detailed switching models system. The LFLS models result in huge cut in simulation time (about 10 times difference) and also the ability to use large time step with the LFLS system and still capture all the information needed. Even though this low frequency large signal (LFLS) models system has wider capabilities than ideal average models system, it still can’t predict all switching phenomena. Therefore, another benefit of this modeling approach is the ability to mix different types of models (low frequency large signal (LFLS) and detailed switching) based on the application study they are used for.