Algorithm for parametric identification of the DC electric drive using the inverse model

Abstract

THE PURPOSE. In modern control systems of various industrial units, the basis for the automation of technological processes is an electric drive. High requirements for the quality of the control process determine the formulation and solution of scientific and practical tasks for the development and creation of new promising control systems for electric drives (EDCS), allowing to maintain the required quality of functioning under the influence of destabilizing factors. These factors can affect the control system in the form of external disturbances and be summed up with the corresponding signals of the closed EDCS, and also be expressed in the form of parametric disturbances. The need arises to create such a control system that will allow for the identification of the parameters of the EDCS by introducing the appropriate identification algorithms into its structure.METHODS. When solving the problem, a search-free gradient method of adaptive identification was used, implemented by means of the MatLab software environment.RESULTS. The paper considers the problem of parametric identification of an electric drive with a DC motor based on the definition of sensitivity functions. Wherein to construct an algorithm for parametric identification, an inverse model of the studied EDCS is used, the quality indicator is the squared discrepancy, and the identifiable parameter is the overall transmission coefficient of the electric drive.CONCLUSION. Modeling in the MatLab software environment showed a high robustness of the developed identification algorithm to parametric disturbances that do not affect the steadystate value of the identified parameter. The proposed method for compensating the moment of resistance electric drive also made it possible to provide a low sensitivity algorithm to external disturbances. The identification of the overall transmission coefficient of a DC electric drive is carried out with an error not exceeding 0.5% in real time under the conditions of the influence of disturbances of various physical nature.