摘要
电力负荷预测是一个较为复杂的过程,由于影响负荷的因素较多,权重向量的选取较为困难,导致负荷预测的准确性较差。通过遗传算法选取合适的权重向量,在范例检索的过程中利用时间序列和组合属性对权重向量和预测结果进行进一步修正,使得负荷预测的精度大大提高,实验结果表明该模型具有有效性和实用性。
Forecasting to electric power load is a complex process. There are many factors affect the load, and it is difficult to choose the weight vector, which lead to the worse veracity. The paper firstly chooses appropriate weight vector by genetic algorithms, then amends the vector and the result of forecasting through making use of time series and attribute- combined during the process of case matching, accordingly the precision is greatly improved. Furthermore, the result of experiments show the model has validity and practicability.
出处
《计算机技术与发展》
2007年第6期197-199,202,共4页
Computer Technology and Development
基金
安徽省教育厅自然科学项目(2006KJ066B)
