基于可加性模糊系统的负荷时间序列预测

Load time series forecasting based on additive fuzzy systems

本文依据可加性模糊系统理论 ,提出了一种新的预测方法 ,利用聚类方法与有监督学习相结合的训练方法 ,提高了系统的函数逼近能力。仿真结果表明 ,系统学习速度快、预测精度高 ,在电力负荷时间序列预测中获得相当满意的结果。

In this paper, we show that additive fuzzy system can be used in the prediction of short\|term load time series. By using optimal cluster algorithm in combination with supervised learning of training system, functional approximation efficiency is improved. Simulations results show that the learning algorithm has the features of quick convergence and high forecasting precision. Fairly satisfactory results can be acquired for power load time series forecasting.