Avelino Juan Gonzalez

On-Line Diagnosis of Turbine-Generators using Artificial Intelligence

An Artificial Intelligence (AI) based, on-line Turbine Generator diagnostic expert system which is is presently under development is described. Utilizing the inputs from a number of sensors, the diagnostic system evaluates the condition of the equipment and communicates appropriate action. This is an improvement over present monitoring systems which simply apply alarm limits to the value of each sensed variable. In comparison to a monitoring system, a diagnostic system identifies equipment and instrument problems earlier with a higher degree of accuracy.

Computer Simulation of Semisynchronous Start of Hydroelectric Pumped Storage Units

A novel method for simulating a semisynchronous start of a pumped storage hydroelectric unit is described. It uses a computer program composed of two sections: an induction start and a synchronous acceleration. The results of the synchronous acceleration portion were extended and verified with other existing, more exact programs. The advantage of this program is that it is written for an HP-85 desk-top computer and can therefore be run very inexpensively. Additionally, it could be run in the field during the commissioning of a unit.

Computer Model Capable of Predicting Arc Instability During Low Current Interrauption

A mathematical model of a low current arc in a gas-blast interrupter is developed which can predict as well as describe current chopping and the resulting transient overvoltage on the connected system. The model is based on Mayrs model, with the exception that it assumes that the time constant and the heat removal rate are not constant, but a function of arc current. A procedure is also included, based on experimental data, which can simulate the point of arc instability and chopping. Simulations are made using this model and comparing it with experimental work to verify its accuracy. This model can be useful in designing circuit interrupters for low current applications.