Abstract:
This thesis represents an investigative analysis of the closed-loop stability of the Unidirectional SEPIC (Single-Ended Primary Inductor Converter) converter, Bidirectional SEPIC converter, and Interleaved SEPIC Converter by implementing Swarm Intelligence Algorithms (SIA) for designing an optimized PID controller. The applicability and compatibility of three Swarm Intelligence Algorithms, which are Firefly Algorithm (FA), Particle Swarm Optimization (PSO), and Ant Colony Optimization for continuous domain (ACOR), are analyzed in optimizing the control mechanism of the power converters. The improvement of performance parameters is observed, such as Percentage of Overshoot (%OS), Rise Time (Tr), Settling Time (Ts), and Peak Amplitude. The outcomes are compared with the help of various fitness functions. The thesis focuses on higher-order SEPIC Converters and its variants (fourth-order). Higher-order converters benefit from smaller ripple currents, easier EMC filtering, and avoiding current spikes owing to resistive losses. The converters were developed using State Space Averaging (SSA), and the transfer function of the converter's open-loop system was determined using MATLAB's system identification toolbox. By using the PID controller, the closed-loop system of the converter is introduced. For the tuning purposes of the PID Controller, the PID Tuner App of MATLAB has been used. Nevertheless, for the better performance of the controller, the algorithms are evaluated in the system through different fitness functions: IAE, ITAE, ISE, and ITSE. MATLAB is used to carry out all the simulations. After analyzing the performances for the case of the Unidirectional SEPIC converter, ACOR-PID (ITSE) is the most optimized controller among all the algorithms based PID controllers in terms of performance parameters. In this case, values of overshoot (1.8603%), settling time (2.3414 sec), and peak amplitude (1.0186) are lower than FA-PID and PSO-PID for each of the error functions. For rise time, the value of ACOR-PID (ITAE) is better (0.3798 sec). Again, for the case of the Bidirectional SEPIC converter, PSO-PID (ITSE) is the most optimized controller among all the algorithms based PID controllers in terms of performance parameters. In this case, values of overshoot (0.2674%) and peak amplitude (1.0027) are lower than FA-PID and ACOR-PID for each of the error functions. For rise time, the value of ACOR-PID (ITAE) is better (0.3798 sec), and for settling time, the value of ACO¬¬R-PID (IAE) is better (0.1134 sec). Furthermore, for the case of the Interleaved SEPIC converter, PSO-PID (ITSE) is the most optimized controller among all the algorithms based PID controllers in terms of performance parameters. In this case, values of overshoot (5.2104%) and peak amplitude (1.0521) are lower than FA-PID and ACOR -PID for each of the error functions. For rise time and settling time, the values of PSO-PID (ITAE) are better (0.1471 sec and 1.3354 sec, respectively). Hence, Swarm Intelligence Algorithm based optimized PID controller provides more optimized results and performs far better than Conventional PID controller for SEPIC converter and its variants.
Description:
Supervised by
Prof. Dr. Md. Ashraful Hoque,
Department of Electrical and Electronic Engineering (EEE),
Islamic University of Technology (IUT),
Board Bazar, Gazipur-1704, Bangladesh.
This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Electrical and Electronic Engineering, 2022.