Вплив тривалості стабілізації параметру в кінці кривої розгону на якість ідентифікації об’єкта керування

This paper examines the impact of parameter stabilization time at the end of the transient response on the accuracy of control system identification. A dataset comprising more than 50 transient response curves of various dynamic systems was analyzed using 15 different identification methods. The study demonstrates that limited stabilization data can result in elevated static errors across all methods, indicating insufficient system dynamics capture. By artificially extending the stabilization phase, the average static error was reduced by 33%, and the coefficient of determination (R²) increased by 1.5%. These results emphasize the importance of including an adequate duration of parameter stabilization in the identification process to enhance model accuracy, while avoiding excessive data that may introduce noise. Future work will focus on evaluating how stabilization time affects the tuning performance of control algorithms.