摘要
影响负荷预测精度的因素众多 ,为了找到负荷值与各种外在因素之间的关系 ,利用粗糙集理论对各条件属性进行属性约简分析 ,在属性约简算法中采用遗传算法进行寻优计算 ,找到与负荷直接相关的因素 ,然后将它作为模糊神经网络的输入矢量进行负荷预测。经仿真分析证明预测精度和速度都得到改善。
There are many factors that influence the accuracy of load forecasting. In order to find the relationship between the load and the outside factors, rough set th eory is used to analyze the condition attributes and to find the relevant factor s to load, these factors are then applied to the fuzzy neural networks as the in put vectors to forecast load. Genetic algorithm is used for optimal searching in attribute reduction. It has been proved that better accuracy and convergence of forecasting are gained by the simulation results.
出处
《电力系统及其自动化学报》
CSCD
2004年第6期60-63,共4页
Proceedings of the CSU-EPSA
关键词
负荷预测
粗糙集
遗传算法
神经网络
load forecasting
rough set
genetic algorithm
n eural network
