Technology and software to determine adequate normalized correlation matrices in the solution of identification problems
| Autor: | Ulkar Sattarova |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Normalization (statistics)
Computer science Diagonal 0211 other engineering and technologies Energy Engineering and Power Technology 02 engineering and technology 01 natural sciences Industrial and Manufacturing Engineering Correlation 010104 statistics & probability Software Management of Technology and Innovation 0101 mathematics Electrical and Electronic Engineering 021101 geological & geomatics engineering Mathematical model business.industry Applied Mathematics Mechanical Engineering Normalized correlation Integral equation Matrix similarity Computer Science Applications Control and Systems Engineering business Algorithm |
| Zdroj: | Eastern-European Journal of Enterprise Technologies. 6:69-76 |
| ISSN: | 1729-4061 1729-3774 |
| DOI: | 10.15587/1729-4061.2017.118265 |
| Popis: | Statistical methods are widely used in solving problems of automatic management of industrial objects, as they enable us to determine the dynamic characteristics during normal operation of objects. The statistical correlation method for determining these dynamic characteristics is based on the solution of an integral equation that includes the correlation functions R XX ( i D t ) and R XY ( i D t ) of the input X ( i D t ) and output Y ( i D t ) signals. It allows one to obtain the dynamic characteristics of an object without disturbing its regular operation mode. However, the application of these methods for constructing mathematical models of real-life industrial objects presents the following certain difficulty. Interferences and noises are imposed upon the useful signal, hindering the calculation of the estimates of their static characteristics. The paper presents one possible option of creating alternative methods and technologies for eliminating the error induced by noise during the formation of correlation matrices. The proposed algorithms allow for reducing these matrices to the similar matrices of useful signals. It is demonstrated in the paper that in the traditional approach, due to the normalization of estimates in the diagonal elements of the correlation matrices, the noise-induced errors disappear, while appearing in the remaining elements. As a result, the expected effect of improving the conditionality from the transition to normalized correlation matrices is not achieved. The technology and software for eliminating this defect are proposed, despite the problems with matrix conditioning. A new software for the rapid formation and analysis of numerous computational experiments confirming the effectiveness of the developed technology is proposed |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|