基于大数据的发电设备监测信息聚合分析

Aggregation analysis of monitoring information for power generation equipment based on large data

针对海量的发电监测大数据,提出一种大数据监测设备信息聚合与评估模型。结合信息聚合的原理,在建立数据时空相关的基础上,将信息聚合模型分为三层:底层数据层实现数据的基础采集和数据的预处理;信息级聚合层实现数据的多维关联分析、特征匹配等;决策级聚合实现信息的挖掘、评估等。最后以风力发电机组2017~2018年采集到的数据作为样本,以时域和频域等参数作为SVM的输入参数,对样本进行测试和识别。结果表明,通过信息聚合和识别,可有效识别不同故障类型,并得到监测数据的关联,大大提升了设备监测的自动化和智能化。

Aiming at the huge data of power generation monitoring,a model of information aggregation and evaluation for large data monitoring equipment is proposed.Based on the principle of information aggregation,the information aggregation model is divided into three layers:the underlying data layer is used to realize the basic data acquisition and data preprocessing;the information level aggregation layer is used to realize the multi-dimensional association analysis and feature matching of data;and the decision level aggregation is used to realize the mining and evaluation of information.Finally,the data collected by wind turbines from 2017 to 2018 are taken as samples,and the time domain and frequency domain parameters are taken as input parameters of SVM to test and identify the samples.The results show that different fault types can be identified effectively by information aggregation and identification,and the correlation of monitoring data can be obtained,which shows certain practical value.