Intelligent identification of power quality disturbances

Abstract

The increasing use of electronics in electrical equipment has made them more vulnerable to power quality disturbances. Customers are becoming more knowledgeable in these issues and demand for more detailed and precise information whenever an unfortunate event occurs. Therefore, power companies are expanding the use of information processing techniques in the attempt to determine the cause and severity of power quality event so that remedy actions can be undertaken swiftly. One such process is to pinpoint the type of disturbance that has occurred. This is achieved through some form of pattern matching where signature patterns from past experiences or simulations are used to match the recorded event. However, there are many uncertainties with the network operating conditions and devices complicating the identification process. Some research works have been done in NTU in this area involving the use of advanced signal processing and fuzzy-logic based pattern matching techniques. This project is to test the proposed method for its robustness in handling the various uncertainties. Sample disturbances will be generated using Matlab and incorporated with various uncertainties. These are then used to test out the proposed method.