NEURAL-NETWORK-BASED ADAPTIVE OUT-OF-STEP PROTECTION STRATEGY FOR ELECTRICAL-POWER SYSTEMS

Application of artificial intelligence to power systems has resulted i n an overall improvement of solutions in many areas. This paper presen ts a new strategy for adaptive out-of-step protection of synchronous g enerators based on neural networks. The paper describes the neural net works architecture adopted as well as the selection of input features for training the neural networks. A feed forward model of the neural n etwork based on the stochastic back-propagation training algorithm has been used to predict the out-of-step condition. Due to power network configuration changes, the performance of the protective relays can va ry. Consequently, a new adaptive out-of-step protection strategy is su ggested in this paper. This depends firstly on detecting the case of t he system through case detection neural networks by some pre-fault loc al measurements at the machine to be protected, and then calculating t he new out-of-step condition through an adaptive routine. The capabili ties of the developed adaptive out-of-step prediction algorithm have b een tested through computer simulation for a typical case study. The r esults of using the proposed algorithm demonstrate the adaptability of the proposed strategy.