Application Study On Adaptive Ralay Protection Based On Intelligent Control

The thesis comprises two parts. The first one includes chapter 2 and chapter 3, which is the research of current protection based on ANN; the second one includes chapter 4 to chapter 6,. which is the study of single-phase reclosing based on intelligent control.When the power system becomes more compl icated, the probabi lily of system problem' s emergence is bigger than before. As the most important one in all kinds of security measures, relay protection has been concerned widely by power system' s personnel. Adaptive relay protection based on intelligent control is a new sort of protection principle which develops with the advancement of power system technology and the progress of intelligent control theory. It is presented to satisfy the higher request for power system relay protection.Chapter 2 expounds basic concept and normal principles of current protection and indicates the defect that current protection is influenced deeply and directly by the electric power network' s connection mode and system' s running mode, so current protection is only applied in the electric power network of 35kV of under 35kV. Chapter 3 introduces the research of current protection based on artificial neural network(ANN) in detail. Making use of the advantages of ANN, such as which can approach arbitrary non-linear function, which can study or realize self-adapt as for unclear of uncertain systems, and the functions of ANN of mode identification and mode classification, according to the characteristic of current protection, the thesis puts forward the current protection based on BP neural network and RBF neural network, which can overcome the fault of traditionalcurrent protection. A lot of simulating calculations have boon done. The results of model test shcMved their feasibility.Chapter 4 introduces the concept of automatic reclosing and single-phase automatic reclosing, explains their advantages and shortcomings. Their common shortcoming is that they act blindly entirely before reclosing and can not distinguish instantaneous fault from permanent fault. Aiming at this shortcoming, the concept and theory of an adaptive single-phase reclosing scenario based on intelligent control was presented in Chapter 5. Its core is that it can distinguish instantaneous fault from permanent fault definitely, avoid the blindness of traditional automatic reclosing and eliminate the harm when reclosing on permanent f'aul t. But in some cases, adaptive reclosing can confound instantaneous fault with permanent fault, so it has fault too. Chapter 6 uses the function of mode identification and mode classification of ANN again and establishes the model of self-phase reclosing based on BP neural network. Through validating a line model, the thesis approves its feasibility.The major creative idea of the thesis is the ANiN model' 's sett i ng up in Chapter 3 and Chapter 6, which differs from the former studies. The input eigenvalues contain all kinds of electric parameters more completely and more perfectly.At the end of the thesis, some important conclusions and future research directions after this research are given out briefly.