Research For Adaptive Distance Protection Of Transmission Line

Adaptive protection, which concerns improving protection performance and system running conditions, is one significant research task of modern relay protection techniques.Distance protection, which works as main protection of transmission line, is an essential protection type. But due to the existence of the fault resistance, the branch current, the zero sequence mutual coupling and so on, the apparent impedances and protection scope of distance protection will change, thereby affecting the performance of distance protection and the safety operation of power system.This paper discusses and analyses several principal points affecting the performance of regular distance protection, such as the fault resistance, the branch current, the zero sequence mutual coupling and so on. Then some concrete application of adaptive concept to distance protection will be analyzed and studied.This paper make transient simulation using ATP about 110kV transmission line, then extract the fault parameter and characteristic element from the result of simulation and generate training samples and testing samples. Using these samples, this paper form an impedance relay scheme based on back propagation neural network under the condition ofone side source and two sides source respectively. After the training of neural network this paper obtain the action characteristic of impedance relay. The tests show that the neural network impedance relay has good function to anti fault resistance and adaptability to variation of system condition. Moreover, it has function to fault phase identification and fault distance measurement.Finally this paper propose the structure diagram of adaptive distance protection based on neural network impedance relay, which describes entire logic procedure from data acquisition unit to trip unit.