中国电机工程学报
中國電機工程學報
중국전궤공정학보
ZHONGGUO DIANJI GONGCHENG XUEBAO
2014年
9期
1446-1453
,共8页
梁永亮%李可军%赵建国%牛林%任敬国
樑永亮%李可軍%趙建國%牛林%任敬國
량영량%리가군%조건국%우림%임경국
在线监测周期%相空间重构%引力搜索优化%相关向量机%变压器油色谱
在線鑑測週期%相空間重構%引力搜索優化%相關嚮量機%變壓器油色譜
재선감측주기%상공간중구%인력수색우화%상관향량궤%변압기유색보
on-line monitoring cycle%phase space reconstruction%gravitation search algorithm%relevance vector machine%transformer chromatography
变压器油色谱在线监测装置的运行成本与其监测周期密切相关,如何根据设备的运行情况调整监测周期,在保证运行效率的同时兼顾经济性,是该文研究的核心问题,在此背景下,提出一种动态调整在线监测周期的方法。首先在理论论述监测周期影响油色谱在线监测装置寿命的基础上,对平稳过程短时监测周期的时间序列数据进行相空间重构,得到最优时延和嵌入维数,并以最优时延作为相对最优监测周期。然后基于引力搜索优化方法和快速相关向量机建立气体浓度自适应预测模型,并设定预警标准,根据预测结果以及其他监测设备监测结果保持或缩短监测周期。仿真计算结果证明:所提气体浓度预测模型具有良好的预测精度;相比较依据产气率注意值,基于气体浓度预测技术的预警方法更适用于短时间间隔、含量较低的气体浓度数据;相比气体含量注意值方法,所提方法能够有效地发现可能出现的异常情况。文中研究提供了一种在不影响监测有效性的前提下,实现油色谱在线监测装置经济效益更大化的可行方法。
變壓器油色譜在線鑑測裝置的運行成本與其鑑測週期密切相關,如何根據設備的運行情況調整鑑測週期,在保證運行效率的同時兼顧經濟性,是該文研究的覈心問題,在此揹景下,提齣一種動態調整在線鑑測週期的方法。首先在理論論述鑑測週期影響油色譜在線鑑測裝置壽命的基礎上,對平穩過程短時鑑測週期的時間序列數據進行相空間重構,得到最優時延和嵌入維數,併以最優時延作為相對最優鑑測週期。然後基于引力搜索優化方法和快速相關嚮量機建立氣體濃度自適應預測模型,併設定預警標準,根據預測結果以及其他鑑測設備鑑測結果保持或縮短鑑測週期。倣真計算結果證明:所提氣體濃度預測模型具有良好的預測精度;相比較依據產氣率註意值,基于氣體濃度預測技術的預警方法更適用于短時間間隔、含量較低的氣體濃度數據;相比氣體含量註意值方法,所提方法能夠有效地髮現可能齣現的異常情況。文中研究提供瞭一種在不影響鑑測有效性的前提下,實現油色譜在線鑑測裝置經濟效益更大化的可行方法。
변압기유색보재선감측장치적운행성본여기감측주기밀절상관,여하근거설비적운행정황조정감측주기,재보증운행효솔적동시겸고경제성,시해문연구적핵심문제,재차배경하,제출일충동태조정재선감측주기적방법。수선재이론논술감측주기영향유색보재선감측장치수명적기출상,대평은과정단시감측주기적시간서렬수거진행상공간중구,득도최우시연화감입유수,병이최우시연작위상대최우감측주기。연후기우인력수색우화방법화쾌속상관향량궤건립기체농도자괄응예측모형,병설정예경표준,근거예측결과이급기타감측설비감측결과보지혹축단감측주기。방진계산결과증명:소제기체농도예측모형구유량호적예측정도;상비교의거산기솔주의치,기우기체농도예측기술적예경방법경괄용우단시간간격、함량교저적기체농도수거;상비기체함량주의치방법,소제방법능구유효지발현가능출현적이상정황。문중연구제공료일충재불영향감측유효성적전제하,실현유색보재선감측장치경제효익경대화적가행방법。
Operation cost of the transformer chromatography online monitoring device and its monitoring cycle are closely related. How to guarantee the monitoring efficiency and economical efficiency by adjusting monitoring cycle is the key issue in this paper. In this context, a dynamic monitoring cycle adjustment strategy for the chromatography online monitoring device was proposed. Based on the theoretical analysis of the impact of the monitoring cycle on transformer service life, phase space reconstruction was carried out on gas content time series data with short time intervals, and optimal time delay obtained was considered as the relative optimal monitoring cycle. Then a self-adaptive gas content forecasting model was built based on gravitation search algorithm and the fast relevance vector machine. The early warning method proposed according to forecasting results along with other monitoring information is used to adjust the monitoring cycle. The numerical results testify that: the forecasting model proposed in this paper has satisfactory performance. Besides, compared with the methods based on the gas production rate alert value and the gas content alert value, the waring method proposed in this paper is more suitable for online data with short time intervals and a low content, and can detect abnormal situation more effectively. The research provides a feasible way of realizing more economic benefits for the transformer chromatography online monitoring device without monitoring effectiveness disturbance.