中文摘要：

为了提高短期电力负荷预测的精度，提出了一种量子行为粒子群优化（Quantum-behaved particle swarm optimization, QPSO）算法和径向基函数神经网络（Radical basis function neural network,RBFNN）相结合的电力负荷短期预测模型。通过K-均值聚类算法确定RBFNN的基函数中心，并用粒子群优化算法优化神经网络权值，在加快RBFNN收敛速度的同时提高预测精度。以实际负荷数据进行预测验证。预测负荷的均方根误差小于0．01．验证了模型的合理性和有效性。

英文摘要：

In order to improve the accuracy of short-term power load forecasting,a forecasting model is proposed by combining quantum-behaved particle swarm optimization （QPSO）algorithm with radical basis function neural network（RBFNN）. The basis function center of the RBFNN is obtained by the K- means clustering algorithm. The neural network weights are optimized by the particle swarm optimization algorithm. The convergence rate of the RBFNN is fastened while the forecasting accuracy is raised. The short-term power load forecasting is verified by real load data,the root mean square error of daily load forecasting is less than 0.01 ,and the rationality and validity of the model are demonstrated.