Implementation of a particle swarm optimization-based controller for a microhydroelectric power plant

• Main Authors: Canapi, Joseph Christian R., Cruz, Elise B., De Ocampo, Domingo T., Jr., Guinto, Patrick William d.V.
• Format: text
• Language: English
• Published: Animo Repository 2007
• Subjects: Hydrodynamics; Swarm intelligence; Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/14296
• Institution: De La Salle University

This thesis implements the Particle Swarm Optimization (PSO) Algorithm in order to control the voltage of a micro hydroelectric power plant model as a requirement for the completion of the degree of Bachelor of Science in Electronics and Communications Engineering. Particle Swarm Optimization is an optimization algorithm based on the flocking behavior of animals in search of food proposed by Dr. Kennedy and Dr. Eberhart. The algorithm was used in the search of the optimum PID constants within a defined solution space. Using the output voltage data acquired by Labjack U12 from the powerplant model, a PC runs a program coded in MATLAB in order to identify the system at the beginning of each operation. The micro hydroelectric power plant model is the identified system, represented by a transfer function in which the PID controller is cascaded to in order to control the plant. Using a PSO-based program, PID constants are then selected, which is used by the PID controller to send an impulse response to the controller interface. The controller interface, which is composed of relays, turns the valve of the power plant model which then controls the output voltage of the plant. Resistors were used to act as loads of the power plant model, and as the loading increased the plant was able to turn the valve appropriately in order to maintain the voltage dictated by the plant's steady state voltage. Cost Function and ANOVA analysis were used in the evaluation of PSO's effectiveness in tuning the PID controller. Cost analysis show that PSO was theoretically able to tune the plant at a lower cost compared to other tuning algorithms such as the Suyama tuning algorithm. In addition, the PID tuned plant had a better step response than that of the uncompensated plant. ANOVA analysis show that the cost resulting from PSO tuning varies significantly from that of the uncompensated plant. Thus, the PSO algorithm is an effective way of tuning a PID controller for the control of a microhydro-electric powerplant.