Simulation Study On Fault Diagnosis Of Transmission Lines Using Data Fusion

With the development of economic and the pursuit of people to higher life quality, it is very important that power system works securely and stably. The system of relay protection and fault diagnosis can ensure power system be in motion smoothly. Facing the world depending on power energy, working personnel are trying every means to solve the problem.Using data fusion technique to realize fault discrimination of transmission lines, this work tried to find a new approach of adaptive relay protection and intelligent fault diagnosis on power system. When transmission lines are at fault, the voltage and current will change violently. This work used data fusion technique based on fuzzy neural network, using parallel multi-sensor to detect the voltage and current as the inputs of neural network, whose distributive information memory and parallel process ability could accomplish multi-sensor data fusion. Finally, at the output points of neural network, the situation assessments were computed out more dependably than the other common approaches, they were given by the possibility of fault.This work also used fuzzy theory in neural network, each detective data was fuzzicated as membership of degree belonging to the "Big", "Middle", "Small" three subsets, converted into mathematical quantized feature that they could be used in neural network, so it could accomplish logical process based on rigorous mathematics theory.This method is so lately that it is still not be used in practice at present, but it can make full use of multi-sensor's information resource and improve the correct possibility of fault diagnosis through increasing the detective dimensions. In addition, it still can give the correct decision at the critical state. In the future, data fusion technique will play an important role in relay protection and fault diagnosis on power system.