Answers are grounded in this paper and cite their sources.
Analysis reveals improved dynamic characteristics in control systems through optimal regulator parameters.
The object of research is the optimal parameters for regulator adjustment and the quality indicators of transient processes. The current problem is that modern technological processes are complex control objects. Therefore, when designing automation systems, the tasks of identifying the control object, calculating the regulator parameters, and their further optimization become particularly important. Optimal settings ensure the highest possible product quality and reduce its cost at a given production volume. Direct determination of controller parameters through experiments on a real object can lead to a loss of finished product quality and damage to raw materials and catalysts. To avoid these risks, the calculation algorithm was implemented in the Maple software package environment. The study developed and tested an algorithm for identifying control objects with different characteristics of transient processes, described by second-order links, taking into account the delay time. Based on the obtained transfer functions of equivalent objects, P-, PI-, and PID-controllers (proportional, proportional-integral, and proportional-integral-derivative) were tuned. The parameters were determined using the triangle method, the method of undamped oscillations (Nicholas–Ziegler), and the proposed algorithm. The results obtained are intended for use in automatic control systems. A comparative analysis of the quality of transient processes in systems tuned using different methods was performed. It was found that the parameters obtained using the new algorithm significantly improve the dynamic characteristics of the system (reduction of overshoot, control time, static and dynamic errors). In addition, an algorithm for searching for controller parameters taking into account the overshoot limitation was proposed, which also showed positive results. The identification error does not exceed 3%, which is acceptable for calculations of this type.
No takes yet. Share an insight, caveat, or question.
Gurin et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: