The measured convergence rates of several algorithms belonging to the class of . gradient methods in function space' are compared. The steepest descent, conjugate gradient and quasi-Newton algorithms, and the effect of different types of linear search routine and initial guess parameters on their convergence rates, arc considered. It is concluded that conjugate gradient methods do offer significant advantages, even for non-linear system models and bounded controls.
The paper discusses the development of a Fuzzy Inference System (ns) based governor control for a pumped storage hydroelectric plant. The First .Hydro Company's plant at Dinorwig in North Wales is the largest of its kind in Europe and is mainly used for frequency control of the U.K. electrical grid. In previous investigations, a detailed model of the plant was developed using MATLAB@/SIMULINK@ and this is now being used to compare FIS governor operation with the Proportional-Integral-Derivative (PID) controller currently used. The paper describes the development of an FIS governor, and shows that its response to a step increase in load is superior to the PID under certain conditions of load. The paper proceeds to discuss the implications of these results in view of the possible practical application of an FIS governor at the Dinorwig plant.
Large-scale integration of wind into power grids can have significant impact on the system stability. This paper presents the Rate Of Change Of Frequency (RoCoF) metric to determine limitations of the system with high levels of wind pene tration. The metric takes into account the system inertia, wind turbine response, and the size of contingencies. The metric is then applied to the Tasmanian power system -a small predominantly hydro power system with a potentially large-scale penetration of wind. The simulation studies are carried out using PSS/E. Case studies are performed with critical contingencies such as the loss of synchronous generation, loss of wind generation and loss of the HVDC interconnector, and their impacts are discussed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.