Abstract: Power load forecast is important in energy management with great influence on operation, controlling and planning of electric power systems. Accurate short-term forecast in electric power load can result in cost saving and better operation conditions. Taking into account the complication and uncertainty of short-term load prediction, this paper proposes a short-term load forecast method using fuzzy-neural network (FNN). The FNN model inherits the strong learning ability of neural networks, and has the capability of fuzzy logic for mapping nonlinear functions and extracting similarity from huge historical data. Experiment results show effectiveness of the proposed FNN model. It is applicable to short-term load forecast in power systems with improved performances.

Key words: short-term power load forecast, fuzzy-neural network, fuzzy-logic