基于需求侧管理的电力用户能耗预测分析

Forecast Analysis of Power Consumption of Power Users Based on Demand Side Management

由于电力供需的巨大不平衡和季节性气候的变化,容易造成巨大浪费,因此基于住宅建筑能耗预测的需求响应将在电力需求侧管理机制中发挥着重要作用。为此,论文提出一种基于用电基础数据对整个地区住宅建筑用电等级进行预测的模型。该模型首先利用数据挖掘技术来发现和总结数据中隐藏的用电模式。其次,将粒子群优化K均值算法应用于聚类分析中,用电量水平由聚类中心划分。最后,提出了一种以支持向量机为基本优化框架的高效分类模型。试验证明,模型的预测性能大大优于传统方法,对合理配置供电容量和提高电网整体质量具有一定的参考价值。

Due to the huge imbalance of electricity supply and demand and the change of seasonal climate,it is easy to cause huge waste,so the demand response based on the prediction of residential building energy consumption will play an important role in the power demand side management mechanism.For this reason,this paper proposes a model to predict the power consumption level of residential buildings in the whole region based on the electricity data.This model firsly uses data mining technology to dis⁃cover and summarize the hidden power consumption patterns in the data.Secondly,the particle swarm optimization K-means algo⁃rithm is applied to cluster analysis,and the power consumption level is divided by the cluster center.Finally,an efficient classifica⁃tion model based on support vector machine is proposed.Tests have proved that the prediction performance of the model is much bet⁃ter than traditional methods,and it has certain reference value for the rational allocation of electric capacity and the improvement of the overall quality of the power grid.