Розв’язання задачі ідентифікації технологічного об’єкта

Abstract. Problem. The identification task is formulated as follows: based on the results of observations of the input and output variables of a system, a model, i.e. a formalized representation of this system, should be built that is optimal in some sense. This shows the connection between the identification task and the above general scheme of establishing regularities based on the results of observations. The identification task is based on modern control theory. To solve it, modern computers are used. The latter, with their high speed and virtually unlimited memory capacity, create the prerequisites for receiving, transmitting and processing large arrays of observations that are necessary for building adequate models of real objects. Goal. Studying the possibilities of solving the problem of identifying technological objects using MATLAB+Simulink. Job tasks: to consider the tasks of identifying technological objects of management; to choose an identification method; identify the electric motor model. Methodology. The research problems should be solved by methods of analysis and synthesis of automatic control systems by the method of least squares. Results. In the work, experimental data on the dependence of the engine idling speed on the excitation current were obtained and solved the identification problem using the MATLAB+SIMULINK package. Originality. Based on the analysis of existing solutions, it was found that the works did not consider the calculation of absolute and relative errors of the results of identifying the object and the model developed using the least squares method. Thus, there is a need to develop a system for identifying a model of a technological object made by the least squares method. To solve the identification problem, an electric motor is chosen as the object of study. Practical value. The paper examines the methods used to identify complex objects. The comparative analysis showed that the most acceptable method is the least squares method, due to simple mathematical calculations in comparison with the other methods. To illustrate the application of this method, two tasks were solved: identification of the model of an electric motor; identification of the model of an alarm system. The studies have shown that model identification can be performed using the MATLAB+Simulink package. The research results obtained on the model differ from the experimental data by no more than 4 %