摘要
飞行实测信号处理中,经常要对具备某种特征的数据进行提取,当数据的特征难以用明确的数学关系描述时,这项工作就只能依赖于人工操作.针对此问题,提出了一种数据的提取方法,其特点在于借鉴了模糊数学的思想,用数学语言对模糊特征进行了明确的描述,从而实现了程序的自动判别和提取.首先给出隶属度分别为0和1时的边界特征,然后构造出特征的隶属函数,再对各采样点进行隶属度计算和二值化处理来判定各采样点是否具备特征,最后经过区间提取和区间合并得到具备该特征的时间区间.经验证,该方法提取数据完整、可靠,已成功应用于某型涡桨发动机指定功率状态数据的提取.
Extracting some kind of feature data is usually needed during processing flight test signal.If the feature is difficult to describe by certain mathematical expression,this work can only be done manually.Aiming at this problem,a kind of method of feature data extraction from flight test signal is put forward.The characteristic of this method use for reference of the theory of fuzzy mathematics and the fuzzy feature is described by mathematical language.Thus the automatic destination and extraction of the program is realized.First of all,the boundary feature is given when grade of membership is 0and 1.Next membership function needs to be determined for judging the degree of data in accordance with some kind of feature.Then according to membership grade calculation of each sampling point and two values processing to determine whether each sampling point has the feature can be judged.Finally time intervals can be obtained after intervals extraction and merging.This method has been successfully applied to the extraction of specified power state data of a certain type of turboprop engine.Result shows that the data extracted is complete and accurate.
出处
《测试技术学报》
2015年第5期379-384,共6页
Journal of Test and Measurement Technology
基金
国家自然科学基金资助项目(50135010)
