您好,欢迎来到中国测试科技资讯平台!

首页> 数字期刊群 >本期导读>基于风电场SCADA系统的云平台设计

基于风电场SCADA系统的云平台设计

3158    2019-10-15

免费

全文售价

作者:梁涛, 李燕超, 许琰, 杨改文

作者单位:河北工业大学人工智能与数据科学学院, 天津 300130


关键词:SCADA;云计算;LS-SVM;并行;风功率预测;遗传算法


摘要:

为充分利用风电场集控中心采集爆炸式增长的海量数据,解决风电场SCADA系统数据存储和处理能力不足的问题,设计基于SCADA系统的云计算平台,实现数据的高效分布式存储以及处理。传统LS-SVM适用于小样本数据,对风电场海量数据的处理比较乏力,针对此问题,采用MapReduce云计算平台实现算法并行化,并用遗传算法对其进行参数寻优。最后利用某实际风电场SCADA系统的历史数据,实现短期风功率预测功能。在数据不断增加的情况下,采用MapReduce云计算平台实现LS-SVM算法并行化训练比单机LS-SVM训练缩短的时间十分显著,可增强SCADA系统的实时性;当数据增加到1.2 GB时,时间缩短一半,而准确度相差0.1%。


Design of cloud platform based on wind farm SCADA system
LIANG Tao, LI Yanchao, XU Yan, YANG Gaiwen
School of Artificial Intelligence, Hebei University of Technology, Tianjin 300130, China
Abstract: In order to make full use of the wind farm centralized control center to collect exploding massive data, solve the problem of insufficient data storage and processing capacity of wind farm SCADA system, a cloud computing platform based on SCADA system is designed to realize efficient distributed storage and processing of data.The traditional LS-SVM is suitable for small sample data, and the processing of massive data of wind farm is relatively weak.To solve this problem,the algorithm is parallelized by MapReduce cloud computing platform,and the parameters are optimized by genetic algorithm.Finally,using the historical data of a real wind farm SCADA system, the short-term wind power prediction function is realized.In the case of increasing data,The LS-SVM algorithm parallelization training using the MapReduce cloud computing platform is significantly shorter than the stand-alone LS-SVM training.the real-time performance of the SCADA system is enhanced. When the data is increased to 1.2 GB, the time is reduced by half, and correct rate differs by 0.1%.
Keywords: SCADA;cloud computing;LS-SVM;parallel;wind power prediction;genetic algorithm
2019, 45(10):114-119  收稿日期: 2018-08-31;收到修改稿日期: 2018-10-09
基金项目: 河北省科技计划项目(16214510D)
作者简介: 梁涛(1975-),男,河北石家庄市人,教授,研究方向为风力发电、大数据、人工智能、自动控制
参考文献
[1] 丛智慧. 基于云平台大数据技术的风电集控系统的设计与应用[J]. 内蒙古科技与经济, 2016(19):61-63
[2] LEE J S, LEE K B. An open-switch fault detection method and tolerance controls based on SVM in a grid-connected T-type rectifier with unity power factor[J]. IEEE Trans.Ind. Electron., 2014, 61(12):7092-7104
[3] ZHANG W Y, ZHANG H G, LIU J H,et al. Weather prediction with multiclass support vector machines in the fault detection of photovoltaic system[J]. IEEE/CAA J. Autom. Sinica, 2017, 4(3):520-525
[4] RAHULAMATHAVAN Y, PHAN R C W, VELURU S, et al. Privacy-preserving multi-class support vector machine for outsourcing the data classification in cloud[J]. IEEE Trans.Dependable Secure Comput, 2014, 11(5):467-479
[5] 杨博, 黄斌. 大数据环境下的支持向量机算法研究[J]. 信息化建设, 2016(3):349
[6] 陈珍, 夏靖波, 杨娟, 等. 基于MapReduce的支持向量机态势评估算法[J]. 计算机应用, 2016, 36(1):133-137
[7] ZHSANG P Y, SHU S, ZHOU M C. An online fault detection model and strategies based on SVM-grid in clouds[J]. IEEE/CAA Journal of Automatica Sinica, 2018, 5(2):445-456
[8] 崔红艳, 曹建芳, 史昊. 一种基于MapReduce的并行PSO-BP神经网络算法[J]. 科技通报, 2017, 33(4):110-115
[9] 罗贤缙, 岳黎明, 甄成刚. 风电场数据中心Hadoop云平台作业调度算法研究[J]. 计算机工程与应用, 2015, 51(15):266-270
[10] 马跃峰, 梁循, 周小平. 一种基于全局代表点的快速最小二乘支持向量机稀疏化算法[J]. 自动化学报, 2017, 43(1):132-141
[11] HONG Y Y, CHEN Y Y. Placement of power quality monitors using enhanced genetic algorithm and wavelet transform[J]. IET Generation, Transmission & Distribution, 2011, 5(4):461-466