Parçacık sürü optimizasyonu (PSO) ile genelleştirilmiş bir ayrı pıd yapısı (FiFNF) kullanılarak pıd kontrolörlerinin otomatik ayarlanması

Proportional-Integral-Derivative (PID) controllers remain a remarkable solution in the field of control systems due to their simplicity, flexibility, and ability to manage a wide variety of processes. Despite the success of PID controllers in various applications, PID tuning remains a persistent challenge during implementation. This tuning process becomes even more complicated when PID controllers are implemented in discrete form, which is necessary for most digital control systems. The discretization of continuous-time PID controllers adds an extra layer of complexity, introducing a wide range of probable discrete PID structures, yielding an inefficient selection process. This thesis introduces a novel approach for automating the tuning of discrete PID controllers by integrating a generalized discrete PID structure, named FiFNF, with Particle Swarm Optimization (PSO). The generalized structure provides flexibility in discretization, allowing automatic selection from 18 different possible discrete PID structures derived from combinations of three discretization methods: Forward Euler Approximation, Backward Euler Approximation, and Trapezoidal Approximation. Experimental simulations in MATLAB and Simulink have been conducted for testing and validating the proposed PID design accordingly. Moreover, a simple comparative analysis is presented between FiFNF and the conventional backward PID configuration, given that both configurations are tuned using PSO.