改进径向函数网格（RBFN）在电力负荷预报中的应用

本文根据电力系统短期负荷预报的特点和径向基函数神经网络的非线性辨识功能，提出了一种负荷预报的新算法－ＲＢＦＮ算法。该算法能够体现负荷的波动性和气候对负荷的影响，收敛速度较快，由于采用丛聚技术调整径向基函数中心，该算法具有较高的预报精度，通过对实际系统的实验表明：可用于提前２４小时的电力系统负荷预报。

Advanced Radial Basis Function Natwork (RBFN) and its Application in Short-Term Load Forecasting

This paper introduces a neural network-Radial Basis Function Network for short-term load forecasting. This algorithm considers the influence of atmospheric temperature on electric load, take a n-means clustering technique to adjust the RBF centers, and use least Squares algorithm for updating the RBFN weights. The examples show that it has more significant advantages in convergence and forecasting accuracy.