摘要
为提高短期负荷预测精度,提出了一种基于遗传算法优化概率神经网络(PNN)的短期预测模型。首先对负荷数据异常值进行辨识与修正,建立PNN短期预测模型,在此基础上引入遗传算法(GA),优化概率神经网络的平滑因子,改善了PNN模型的性能,优化后的PNN短期预测模型预测精度得到明显的提高。实例预测结果证实了该方法的有效性。
In order to improve the prediction accuracy of short-term load forecasting, the method based on probability neural network （PNN） optimized by genetic algorithm （GA） is proposed in this paper. After data identification and correction of load, the PNN forecasting model is established, followed by the introduction of genetic algorithm optimize the smoothing parameters of PNN to improve the performance of the PNN and the optimized PNN short-term load forecasting model accuracy has been improved obviously. The effectiveness of the proposed method is verified by examples.
作者
彭钟华
PENG Zhong-hua(Shenzhen Pumped Storage Power Station Company, Shenzhen 518115, Guangdong, China)
出处
《电气开关》
2017年第1期49-51,56,共4页
Electric Switchgear
关键词
概率神经网络
平滑因子
遗传算法
短期负荷预测
probability neural network
smoothing parameter
genetic algorithm
short-term load forecasting
