基于LSSVM的负控数据填补方法

A LSSVM Based Load Monitoring Data Filling Method

为解决山区小水电的负荷监控终端因日常维护困难、通信通道故障多而导致的数据缺失问题,提出一种基于最小二乘支持向量机(LSSVM)的负控数据填补方法。该方法利用LSSVM回归理论拟合训练样本,可实现具有同趋性多源数据的批量化填补。案例分析结果表明,该方法具有较好的数据还原效果,具有较强的工程实用和推广价值。

In order to solve data missing problem of load monitoring terminal of small hydropower in mountain area which is caused by difficult daily maintenance and frequent communication channel faults, a least square support vector machine （LSSVM） based load monitoring data filling method is proposed. The method fits the training samples by using LSSVM regression theory which can realize the batch filling of multiple source data with same trend. Cases analysis results show that the method has better data recovery effects and stronger practical and promotional value of project.