Transmission lines fault location estimation based on artificial neural networks and power quality monitoring data
Fault location estimation is a very important question in electric power system, in order to isolate the fault as soon as possible, and to recover the system with minimal interruptions. In that way, electric equipment is less stressed, and buyers more satisfied. Electric power lines are exposed to environment and probability of line failure is generally higher than other system element failure. Current electric power systems are equipped with high sampling rate power quality meters that are installed in the places of common coupling with distribution systems or high voltage consumers. Data obtained by these power quality meters, especially the voltage and current harmonics present a valuable information about system behavior, even in the faulty conditions. In this paper fault location and fault resistance is estimated by using a combination of artificial neural networks and voltage and current harmonics measured by power quality meters installed only in important system busbars. Results obtained from the real 110 kV transmission system show that a proposed algorithm can be used successfully in fault location and fault resistance estimation in one part of the electric power system. This paper makes a contribution to the existing body of knowledge by developing and testing a new method whose application represents a natural and a feasible upgrade using the existing measurement and communication equipment.