基于神经网络的电网建设需求智能估算模型研究

针对各省级电网建设需求的复杂非线性，非平稳性特征以及地域间差异性特点，提出一种基于系统聚类分析（Hierarchical Cluster Analysis，HCA）与径向基（Radial Basis Function，RBF）神经网络的省级电网建设需求估算模型。该模型利用系统聚类分析方法对省级电网建设需求进行聚类分析，将分析得到的不同类别省份数据作为径向基神经网络的训练样本，挖掘各类别数据中所蕴含的规律，建立了考虑数据差异化特质的预测模型。运用国家电网公司所辖24家省级电网所在地区的相关历史数据，开展了实证研究，算例结果证实了本文所建预测模型的实用性和有效性。

Research on intelligent estimation model of power grid construction demand based on neural networ

According to characteristics of complex nonlinearity,nonstationarity and regional differences of provincial power grid construction demand,a provincial power grid construction demand estimation model based on system hierarchical cluster analysis and radial basis function neural network is proposed.The model uses system cluster analysis method to cluster construction demand of provincial power grid,taking provincial data of different categories as the training samples of radial basis function neural network,excavates the laws contained in each category of data,and establishes a prediction model considering the characteristics of data differentiation.Using the relevant historical data of 24 provincial power grids under jurisdiction of State Grid Corporation of China,an empirical study is carried out.The example results confirm the practicability and effectiveness of the prediction model established in this paper.