基于改进模糊神经网络的电力系统短期负荷预测

Short-term load forecasting in power system based on improved fuzzy neuro net

提出了基于改进聚类算法的模糊神经网络的短期负荷预测方法。首先,利用改进聚类算法确定模糊神经网络的结构,然后利用混合学习算法训练该网络的前件和结论参数,最后向训练好的模糊神经网络输入相关的影响因素数据进行预测。预测结果显示,改进的模糊神经网络可以获得较高的预测精度,所以有更好的使用价值。

An improved fuzzy neuro net based on T-S model is presented in this paper.Firstly,the structure of fuzzy neuro net（FNN）is decided by improved clustering,then the primise parameters and consequent parameters of FNN are trained by hybrid learning algorithm,finally related data are input into the trained FNN to forecast electricity load.The results show that this method can obtain higher forecasting accuracy,so it has more value.