ASVT51 method and the site created Tibabishev Valeriy Nikolaiy. (VN).
The author of a new identification method ASVT61 is Tibabishev Sergei Valeriy. (SV).
The first work that addresses the problem of identification, was published (VN) as an inventor's certificate to a method for determining the phase frequency response of facilities management in 1968. This method was used to obtain a mathematical model of a coal miner. Starting from 1985 to work on the problem of identification attached (SV). Execution of the work to be used for the accelerated acceptance tests of communications equipment. The total time, given to us in the task of identification of mathematical models of different facilities management, exceeded a half century. For a long time we have worked with specialists of the department "Automatic Control Systems" at the Moscow Higher Technical School. Bauman. If during the long years of work and we will be able to bring a small grain of sand in the construction of the temple of science, it is because we had the opportunity to work under the guidance of professors Bauman Vladimir Solodovnikov, Alexei Dmitriev and Vladislav Biryukov.
In the period from 1973 by 1976 the work was performed (VN), led by Professor V.V. Solodovnikov dedicated to identifying the dynamic characteristics objects of industrial control in the class of non-parametric linear time-invariant models. The developed algorithm allows to find quasi-optimal regularized solution of the problem of identification according to a passive experiment (on the signals observed in the course of normal operation of the facility management) in the absence of a priori information about the spectrum of the exact solution and the spectrum of noise, when the ratio of noise / signal frequency identification did not exceed three units. In the future, the algorithm has been updated (SV) and the restriction on the ultimate value of the ratio noise / signal was removed. Created versions of the algorithm to identify the as a stand-alone and closed external feedback.
The next joint work was done under the guidance of Professor A.N. Dmitriev devoted to identification of parameters of the mathematical model, the dynamics of which can only be described at the same time, using ordinary differential equations , and partial differential equations. As such was a slowly rotating spacecraft with oscillating spatial elastic elements. Has developed a mathematical model of the structure is determined in the normal form of Cauchy, which has minimal dimension. The model and identification algorithms used to create the system of precision angular stabilization of large orbital telescope.
Regular opponent of our work was Professor V.F. Biryukov. We are very grateful for the constructive criticism and valuable comments. V.F. Biryukov significantly to change our views on the classical foundations used in the theory and practice of object identity management. In part, our understanding of these issues is reflected in the report "The paradoxes of the practical application of some theoretical principles for solving the problem of identification" made at the "XXVII academic readings on space memory of Academician S.P. Korolev and other prominent Russian scientists - pioneers of space exploration." (Moscow, 2002).
Now ready for practical use a fundamentally new method for identifying the dynamic characteristics of multiply and multidimensional control objects according passive experiment in the class of linear stationary models - ASVT51. The method has high noise immunity, as shown by tests on models. Interference power, distorting the input and output signals can exceed the power of the useful signals several times. With meager input spectrum concentrated in the low frequency region, the method allows to determine the identity ASVT51 order and coefficients of the ordinary differential equation, the value of the transport delay, all the frequency response at any given frequency range.
To automate the process of identifying a program of training data PODAN51, which allows you to make the source data to be sent to us via e-mail and browse from e-mail the results of identification. The results of the identification procedure and are the coefficients of the ordinary differential equations, all frequency characteristics that are presented in tabular and graphical form. PODAN51 program allows to count the frequency characteristics in a given frequency range. You can download PODAN51, if you accept the agreement.
Work is underway to generalize the method to identify objects ASVT51 control in the class of non-stationary models.
ASVT51 planned to generalize the method to identify the management objects in the class of nonlinear models.
ASVT51 identification method used in those cases where all input signals of multiply and multivariate control object can be identified independent components. If this is not possible, then in such cases, we recommend to solve the problem of identification by "gentle experiment" ASVT61. The name of the method suggested Professor Bauman A. N. Dmitriev. Currently, work is underway to develop a method for identifying ASVT61. This method is active due to the use of a special test signal, but the feature of the method lies in the fact that it can be applied to normally functioning facility management. The test signal is added to the current control signal, thus, in most cases, the value of the level of the test signal can be set such that the parameters of object motion control will not exceed the specified error margin management, the latter is characteristic only for passive authentication methods. It was planned that (SV) will create a website ASVT61 a gentle method of experiment. Recently (SV) became interested in other scientific problem and ASVT61 site was not created.
Currently (VN) busy with two objectives: first, to give the tried and tested software products in a form that allows their utilization by the experts, and have created this site, and secondly, are working on new methods of identification, which according to the passive experiment will address the task of identifying a broad sense, ie, define a class of models (linear, nonlinear, stationary, non-stationary and their possible combinations in the presence of strong interference).
Part of our work have been published in the journals Proceedings of the University of the USSR'' INSTRUMENT'', Proceedings of the MSTU. Bauman, Bulletin of Inventions of the USSR, in magazines SAP, presented at several conferences.
Back to main page