A least squares method for sums of functions satisfying the differential equations with polynomial coefficients



метод наименьших квадратов, дифференциальные уравнения, линейные задачи, полиномиальные коэффициенты, разделение сигналов, безусловная минимизация функций


We propose a linear algorithm for determining the parameters of two functions on the basis of their linear combination. These functions must satisfy first-order differential equations with polynomial coefficients, whereas the parameters to be found are the coefficients of these polynomials. The algorithm is based on the least squares method and consists of sequential solution of the following two linear problems: determining the coefficients of polynomial terms in the differential equation satisfied by a linear combination of two given functions and determining the function parameters with the use of these polynomial coefficients. Numerical results obtained according to the above scheme confirm good performance of our method under weak normal noise (with dispersion less than 3 per cent)

Author Biographies

O.I. Berngardt

A.L. Voronov


  1. Камке Э. Справочник по обыкновенным дифференциальным уравнениям. М.: Наука, 1976.
  2. Дэннис Дж., Шнабель Р. Численные методы безусловной оптимизации и решения нелинейных уравнений. М.: Мир, 1988.
  3. Корн Г., Корн Т. Справочник по математике для научных работников и инженеров. М.: Наука, 1974.
  4. Petersson J., Holmstrom K. Methods for parameter estimation in exponential sums // Research Reports in Matehematics/Applied Mathematics. Technical Report IMa-TOM-1997-5. Petersson-Holmstrom: Mälarden University, 1997.
  5. Feldman A., Whitt W. Fitting mixtures of exponentials to long-tail distributionas to analyze network performance models // AT&T Laboratory-Research. Presented at IEEE INFOCOM’97. Kobe (Japan), 1997.
  6. Chocholaty P. A method of inversion of the Laplace transform // Math. Slov. 1992. 42, N 2. 239-246.
  7. Chem T., Rehg J. A multiple hypothesis approach to figure tracking // Cambridge Research Laboratory. Technical Reports Series. CRL 98/8. Cambridge, 1998.
  8. Laidlaw D. Material classification of magnetic resonance volume data. Thesis for the Master of Science Degree. California Institute of Technology. Pasadena, 1992.
  9. Dennis J.E., Schnabel R.B. Numerical methods for unconstrained optimization and nonlinear equations. Englewood Clifs: Prentice-Hall, 1983.
  10. Куликов Н.К., Багаутдинов Г.Н. Обыкновенные дифференциальные уравнения. Решение дифференциальных уравнений на основе функций с гибкой структурой. Алма-Ата, 1973.
  11. Уилкинсон Дж., Райнш К. Справочник алгоритмов на языке АЛГОЛ. Линейная алгебра. М.: Машиностроение, 1976.



How to Cite

Бернгардт О.И., Воронов А.Л. A Least Squares Method for Sums of Functions Satisfying the Differential Equations With Polynomial Coefficients // Numerical Methods and Programming (Vychislitel’nye Metody i Programmirovanie). 2003. 4. 167-171



Section 1. Numerical methods and applications