In the paper, a fusion method of rough set and neural network for fault is put forward and used in generator fault diagnosis. At first, rough set theory is utilized to reduce attribute of diagnosis system. Set in accordance with the practical needs, optimized decision attribute set acts as the input of artificial neural network used for fault diagnosis, which has been used for Fengman hydroelectric power station and testified the feasibility of integration of rough set and neural network. Given enough data, this method could be popularized to other generators.
展开▼