论文部分内容阅读
如何实现准确有效的电力需求预测,是电力部门需要解决的重点及难点工作之一。本文以北京市中长期电力需求预测为例,将文本挖掘技术引入到电力需求预测领域,通过偏差分析的方法,利用REP-Tree决策树技术和RBF神经网络模型对影响北京市中长期电力需求的各种文本类因素展开文本分析挖掘,建立了基于文本挖掘的电力需求预测模型,该模型不仅可以提高预测精度,增强预测结果的稳定性,还突破了以往预测模型只能输入具体数值数据的局限,将各种文本因素对电力需求影响的程度判断转化为具体的预测数值,实现预测方法上的优化。
How to achieve accurate and effective power demand forecasting is one of the key and difficult tasks to be solved by the power sector. This paper takes the medium- and long-term power demand forecast in Beijing as an example to introduce the text mining technology into the power demand forecasting field. By using the method of deviation analysis and the REP-Tree decision tree and RBF neural network model, Various types of textual factors to start text analysis and mining, the paper establishes a power demand forecasting model based on text mining. This model can not only improve forecasting accuracy and stability of forecasting results, but also break through the limitations of past forecasting models that only input specific numerical data , Will be a variety of textual factors on the impact of power demand judgments into specific predictive value, to achieve forecasting method optimization.