基于LSSVM的机床设备电气负荷预测

State of Texas,USA:The University of Texas at Arlington,2015.Machine Tool Equipment Electric Load Forecasting Based on LSSVM

针对机床设备中电气负荷波动性大的问题,在研究超短期负荷基本预测方法的基础上,提出利用聚类分析方法改进最小二乘支持向量机方法。该方法利用K—means方法将前期负荷序列与同期序列进行对比,并将对比结果引入至预测样本的构建中,同时利用该方法对处理后的样本进行建模训练,从而形成短期预测模型。对比分析表明,经该方法改进后的预测数据具有准确性高,适应性强的优点。

On the basis of studying the basic prediction method of ultra-short-term load,in view of the fluctuations of machine tool equipment,this paper proposes a method of improving least square support vector machine by cluster analysis.It uses the K-means method to compare the pre-load sequence with the corresponding sequence,and putting the results of the comparison into the construction of the prediction sample.At the same time,this method is used to build the processed samples to form a short-term prediction model.The comparative analysis shows that the improved prediction method has the advantages of high accuracy and adaptability.