论文:2016,Vol:34,Issue(1):166-175
引用本文:
梁红. 差值定理在离散数据一阶导数解算中的应用[J]. 西北工业大学学报
Liang Hong. Applying Difference Theorem to Calculating First Order Derivatives of Discrete Data[J]. Northwestern polytechnical university

差值定理在离散数据一阶导数解算中的应用
梁红
中国人民解放军91550部队94分队, 辽宁 大连 116023
摘要:
针对差值定理在离散数据一阶导数解算中易受测量误差影响而导致一些测量数据难以获得解算结果的问题,对差值定理的适用条件进行了解析,更正了已有文献中的错误,并对其应用方法进行了分析和推导,提出了基于极值点判别原则下差值定理与最小二乘算法相结合并对三次拟合多项式的一次项系数和二次项系数进行调整的一种新的离散数据一阶导数解算方法,给出了等间隔采样条件下的计算公式。仿真数据和实测数据验证结果表明,新算法能够对测量序列不包括端点在内的所有数据的一阶导数进行有效解算,解算结果不受测量误差限的影响,且解算精度总体上优于不进行多项式系数调整的情况,使差值定理能够更好地进行工程化应用,可显著改善测量序列端点附近和剧烈变化段一阶导数解算精度差的状况。
关键词:    差值定理    微分    一阶导数    数字滤波    截断误差   
Applying Difference Theorem to Calculating First Order Derivatives of Discrete Data
Liang Hong
PLA Unit 91550, Dalian 116023, China
Abstract:
When the Difference Theorem is applied to calculating first order derivatives of discrete data, the measurement error can lead to some of the data having no result. To resolve this problem, application conditions of the Difference Theorem are analyzed, and the inaccuracies about them in some literature are corrected. Through analyzing and deducing the application method, a new algorithm on first order derivatives of discrete data is put forward; it combines the Difference Theorem with least squares algorithm and adjusts item coefficient and binomial coefficient of third-order fit polynomial for extreme point identifying, and the calculation-formula is given on the premise that measurement data are at equal intervals. Verification of the algorithm is made with simulation and measurement; the results indicate that: (1)this new algorithm can calculate effectively first order derivatives of all data of the measurement sequence, exclusive of the endpoints; (2)the measurement error bounds do not affect the results; (3)the calculated precision is better in general than that of the case in which polynomial coefficients are not adjusted. These enable the Difference Theorem to improve significantly calculation precision of the first order derivatives of the points in the interval where the data change abruptly or near the endpoints of measurement sequence.
Key words:    algorithms    calculations    efficiency    errors    functions    Kalman filters    least squares approximations    linear regression    measurements    polynomials    schematic diagrams    statistics    Difference Theorem    differential coefficient    digital filtering    first order derivative    truncation error   
收稿日期: 2015-04-18     修回日期:
DOI:
通讯作者:     Email:
作者简介: 梁红(1971-),女,91550部队高级工程师,主要从事试验数据处理方法研究。
相关功能
PDF(1088KB) Free
打印本文
把本文推荐给朋友
作者相关文章
梁红  在本刊中的所有文章

参考文献:
[1] 王业桂,蔡其发,黄思训. 一种气象观测数据求导的新方法[J]. 物理学报, 2010, 59(6): 4359-4368 Wang Yegui, Cai Qifa, Huang Sixun. A New Method for Calculating the Derivation of Meteorological Observational Data[J]. Acta Physica Sinica, 2010, 59(6): 4359-4368 (in Chinese)
[2] 孙中豪,杜娟,王子龙. 基于白噪声正交多项式滤波的GPS测速方法分析[J]. 测绘地理信息,2013,38(5):21-24 Sun Zhonghao, Du Juan, Wang Zilong. Analysis of GPS Velocimetry Method Based on White Noise Orthogonal Polynomial Filtering[J]. Journal of Geomatics, 2013, 38(5): 21-24 (in Chinese)
[3] 吕游. 基于过程数据的建模方法研究及应用[D]. 北京: 华北电力大学, 2014 Lü You. The Research and Application of Data-Based Process Modeling Method[D]. Beijing, North China Electric Power University, 2014 (in Chinese)
[4] 刘继军. 不适定问题的正则化方法及应用[M]. 北京:科学出版社,2005
[5] 刘也,朱炬波,梁甸农. 递推样条滤波的工程化应用研究[J]. 宇航学报,2010,31(12): 2794-2800 Liu Ye, Zhu Jubo, Liang Diannong. Research of Recursive Spline Filter for Engineering Applications[J]. Journal of Astronautics,2010,31(12):2794-2800 (in Chinese)
[6] 袁莉,刘宏伟,保铮. 雷达高分辨距离像分类器的参数自适应学习算法[J]. 电子与信息学报, 2008,30(1):198-202 Yuan Li, Liu Hongwei, Bao Zheng. Adaptive learning of Classifier Parameters for Radar High Range Resolution Profiles Recognition[J]. Journal of Electronics & Information Technology, 2008,30(1):198-202 (in Chinese)
[7] 梁红. 利用差值定理降低飞行器速度和加速度拟合的截断误差[J]. 飞行器测控学报,2005,24(3):51-54 Liang Hong. Using Error Theorem to Reduce the Truncation Error Produced in Calculating the Velocity and Acceleration of Spacecraft[J]. Journal of Spacecraft TT & C Technology, 2005, 24(3): 51-54 (in Chinese)