电网技术
電網技術
전망기술
POWER SYSTEM TECHNOLOGY
2014年
9期
2479-2483
,共5页
通用动态负荷模型%电压等效阶跃点%聚类%小波
通用動態負荷模型%電壓等效階躍點%聚類%小波
통용동태부하모형%전압등효계약점%취류%소파
general dynamic load model%voltage equivalent step point%clustering%wavelets
如何确定渐变电压信号等效阶跃点位置和采取合理的信号预处理方法对通用动态负荷模型参数辨识有重要影响。针对动态负荷建模试验的需要,利用等效电压阶跃来代替电压渐变信号,把聚类分析引入等效电压阶跃点检测中。在一定时间区间内考虑电压数值的分布特性,结合小波去噪进行负荷信号预处理,基于MATLAB平台,借助GUI(图形用户界面)开发手段,实现了通用动态有功负荷模型的参数辨识。实验结果表明:采用聚类分析方法,可以有效实现经均值滤波和小波去噪处理后的渐变电压信号的阶跃点检测;利用bior3.7小波4层分解去噪的信号进行参数辨识,结果稳定性最好。该方法适用于由于手动改变变压器分接头位置等引起的电压渐变条件下具有功率恢复特性的负荷建模和离线分析。
如何確定漸變電壓信號等效階躍點位置和採取閤理的信號預處理方法對通用動態負荷模型參數辨識有重要影響。針對動態負荷建模試驗的需要,利用等效電壓階躍來代替電壓漸變信號,把聚類分析引入等效電壓階躍點檢測中。在一定時間區間內攷慮電壓數值的分佈特性,結閤小波去譟進行負荷信號預處理,基于MATLAB平檯,藉助GUI(圖形用戶界麵)開髮手段,實現瞭通用動態有功負荷模型的參數辨識。實驗結果錶明:採用聚類分析方法,可以有效實現經均值濾波和小波去譟處理後的漸變電壓信號的階躍點檢測;利用bior3.7小波4層分解去譟的信號進行參數辨識,結果穩定性最好。該方法適用于由于手動改變變壓器分接頭位置等引起的電壓漸變條件下具有功率恢複特性的負荷建模和離線分析。
여하학정점변전압신호등효계약점위치화채취합리적신호예처리방법대통용동태부하모형삼수변식유중요영향。침대동태부하건모시험적수요,이용등효전압계약래대체전압점변신호,파취류분석인입등효전압계약점검측중。재일정시간구간내고필전압수치적분포특성,결합소파거조진행부하신호예처리,기우MATLAB평태,차조GUI(도형용호계면)개발수단,실현료통용동태유공부하모형적삼수변식。실험결과표명:채용취류분석방법,가이유효실현경균치려파화소파거조처리후적점변전압신호적계약점검측;이용bior3.7소파4층분해거조적신호진행삼수변식,결과은정성최호。해방법괄용우유우수동개변변압기분접두위치등인기적전압점변조건하구유공솔회복특성적부하건모화리선분석。
It is of significance for parameter identification of general dynamic load model to determine the position of equivalent step point in gradually-varying voltage signal and adopt reasonable signal preprocessing method. To meet the requirement of the dynamic load modeling experiment, the equivalent voltage step is adopted to take the place of gradually varied voltage signal, and the clustering analysis is led into the detection of equivalent voltage step point. Considering distribution characteristic of numerical value of voltage in a certain time interval and combining with wavelet denoising, the load signal is preprocessed;based on the platform of MATLAB and by means of development tool of graphics user interface (GUI) the parameter identification of general dynamic active load model is implemented. Experimental results show that, with the clustering method, the gradually-varying voltage step point can be effectively detected on the basis of the signal processed by average filtering and wavelet denoising;and utilizing the signal denoised by bior3.7 wavelet with the 4-th level decomposition for parameter identification, the obtained result is of the highest stability. The proposed method is suitable for load modeling with power recovery characteristic and off-line analysis under the condition of gradually-varied voltage due to manually changing the position of transformer tap and so on.