Springer, 2003. — 204 p.
In this book the well-known methods of analysis of noisy signals are revised for the first time. A computer technology of analysis of noise as an information carrier is developed. This technology allows us to determine the estimate of the noise variance, its correlation function, its distribution law, as well as the coefficient of correlation and cross correlation function between the noise and the signal. For the case when classical conditions are not fulfilled the theoretical fundamentals of robust statistical analysis are developed. That makes it possible to improve the estimate of correlation and spectral characteristics, correlation matrix stipulation, and identification adequacy considerably. All of this allows us to realize the solution of a large number of the most important problems, which was impossible in the framework of classical algorithms. The results of computer experiments, examples of solving various problems of forecasting, diagnostics, identification in fields of seismology, petroleum prospecting, oil-gas production, and oil chemistry are given.
This book is intended for specialists who work in various fields of science and technology on applied problems on personal computers.
Needs in Development of Statistical Analysis Technology
Robust Methods for Determining Estimates of Statistical Characteristics
Robust Technology of Correlation Analysis
Robust Technology for Improving Correlation Matrix Stipulation and Identification Adequacy
Robust Technology of Harmonic Analysis
Position-Binary Technology of Statistical Analysis of Cyclic Processes and Noises
Algorithms and Technology of Analysis of Interference as Information Carrier
Application of Robust Technology for Solving Problems of Control, Diagnostics, and Forecasting