АЛГОРИТМ ПРЕДВАРИТЕЛЬНОЙ ИДЕНТИФИКАЦИИ ПАРАМЕТРОВ СХЕМЫ ЗАМЕЩЕНИЯ АСИНХРОННОЙ МАШИНЫ С КОРОТКОЗАМКНУТЫМ РОТОРОМ ПО КРИВОЙ ЗАТУХАНИЯ ФАЗНОГО ТОКА

Abstract

Relevance The construction of the most popular and widespread, according to the author, a field oriented control system for squirrel-cage (SC) induction machine (IM) is based on the use of a linear mathematical model IM, the parameters of which are represented by a T-shaped equivalent circuit without taking into account magnetic core loses. The most domestic frequency converters (FC) do not have a built-in universal option for estimation of IM equivalent circuit parameters under the condition of the rotor non-rotatable, which in turn is the standard for the most advanced foreign manufacturers. Thus, the construction of offline identification algorithms is especially important in the development of software for universal domestically produced frequency converters, capable of working with different AM models, which increases their competitiveness. Aim of research To develop an algorithm for off-line identification of non-rotating SC IM equivalent circuit parameters by a stator falling current curve using the differential evolution method based on the obtained analytical mathematical model, and also to check the correctness of the obtained parameter estimates using analysis of regression residuals. Research methods In this study, theoretical and experimental research methods were used. The theoretical methods include: the theory of electric drive, the theory of automatic control systems, the theory of electrical machines, the theory of differential equations, as well as methods for compiling and solving systems of linear and non-linear algebraic equations, direct and inverse Laplace transform, optimization methods. Experimental studies were carried out on a hardware-software complex, where regression analysis methods were used to evaluate the accuracy of the results. Results An adaptive regression mathematical model has been developed that describes in an analytical form the process of SC non-rotating IM stator current falling. Based on the developed adaptive regression mathematical model, an algorithm of induction motor equivalent circuit parameters estimation was formed by the method of differential evolution. Using the analysis of regression residuals, confirmed the correctness of estimated IM equivalent circuit parameters obtained using the developed preliminary identification algorithm based on the experimental falling current curves. Keywords: equivalent circuit, induction machine, analysis of regression residuals, offline identification, differential evolution, optimization