系统工程与电子技术
繫統工程與電子技術
계통공정여전자기술
SYSTEMS ENGINEERING AND ELECTRONICS
2014年
9期
1707-1711
,共5页
系统误差%相位差测量%全相位离散傅里叶变换%迭代延时%补偿法
繫統誤差%相位差測量%全相位離散傅裏葉變換%迭代延時%補償法
계통오차%상위차측량%전상위리산부리협변환%질대연시%보상법
systematic error%measurement of phase difference%all-phase discrete Fourier transform (apDFT)%delay-compensation method based on iteration
分析了负频率分量对全相位离散傅里叶变换（all-phase discrete Fourier transform，apDFT）测量相位差的性能的影响，提出了迭代延时补偿法来抑制负频率分量带来的系统误差，讨论了加窗对迭代延时补偿法系统误差抑制性能的影响。当采用 apDFT 算法测量相位差时，负频率分量引起度量级的系统误差。迭代延时补偿法与 apDFT 算法的结合，能够在离散傅里叶变换（discrete Fourier transform，DFT）运算点数为1024时，将最大系统误差在全频段范围内降至0．01°量级，且 DFT 运算点数越大，抑制效果越好。对原始数据加余弦窗，能够进一步改善迭代延时补偿法的系统误差抑制性能及容差性。
分析瞭負頻率分量對全相位離散傅裏葉變換（all-phase discrete Fourier transform，apDFT）測量相位差的性能的影響，提齣瞭迭代延時補償法來抑製負頻率分量帶來的繫統誤差，討論瞭加窗對迭代延時補償法繫統誤差抑製性能的影響。噹採用 apDFT 算法測量相位差時，負頻率分量引起度量級的繫統誤差。迭代延時補償法與 apDFT 算法的結閤，能夠在離散傅裏葉變換（discrete Fourier transform，DFT）運算點數為1024時，將最大繫統誤差在全頻段範圍內降至0．01°量級，且 DFT 運算點數越大，抑製效果越好。對原始數據加餘絃窗，能夠進一步改善迭代延時補償法的繫統誤差抑製性能及容差性。
분석료부빈솔분량대전상위리산부리협변환（all-phase discrete Fourier transform，apDFT）측량상위차적성능적영향，제출료질대연시보상법래억제부빈솔분량대래적계통오차，토론료가창대질대연시보상법계통오차억제성능적영향。당채용 apDFT 산법측량상위차시，부빈솔분량인기도량급적계통오차。질대연시보상법여 apDFT 산법적결합，능구재리산부리협변환（discrete Fourier transform，DFT）운산점수위1024시，장최대계통오차재전빈단범위내강지0．01°량급，차 DFT 운산점수월대，억제효과월호。대원시수거가여현창，능구진일보개선질대연시보상법적계통오차억제성능급용차성。
The influence of negative frequency component on the measurement of phase difference is ana-lyzed when the all-phase discrete Fourier transform (apDFT)algorithm is utilized.A delay-compensation meth-od base on iteration is put forward to reduce the systematic error caused by the negative frequency component, and its performance in reducing the systematic error is discussed when the window function is applied.When the apDFT algorithm is used to measure phase difference,the negative frequency component can cause a systematic error of several degrees.The delay-compensation method based on iteration combined with the apDFT algorithm can reduce the maximal systematic error to 0.01°level in the whole bandwidth when the number of points for discrete Fourier transform (DFT)is 1 024.Besides,the larger the number of points for DFT is,the better of the performance will be.When the original data is windowed by a cosine function,the performances of the delay-compen-sation method based on iteration are improved in the reduction of systematic error and the tolerance of noise.
