Investigations on PID Controller Tuning Employing Soft Computing Techniques for Boost DC to DC Converter

Abstract

The power engineering research mainly focuses in the area of boost converter for various high power applications. The power from the supply is not directly given to the load because of the reasons, such as reduced capability of the load to draw maximum power at certain instant, and the possibility of load damage due to direct connection, and hence a Direct Current newline- Direct Current (DC-DC) converter is needed as an interface between the supply and the load. Boost converter is a type of DC-DC converter that has an output voltage magnitude greater than the input voltage is used as an interface between the supply and the load. It has the most desirable features such as high eand#64259;ciency, high power density, high reliability and reduced Electromagnetic Interference (EMI). The duty cycle of the boost converter is varied using Proportional Integral Derivative (PID) controller. In order to overcome the drawbacks associated with the conventional controllers while used in industries, the PID controller is preferred. Designing and tuning a PID controller in boost converter for speed control appears to be conceptually easy, but hard in practice if enhanced dynamic response is also needed to be achieved. The research concentrated on four main contributions for the optimal tuning of the parameters of PID controller, such as Queen Bee Genetic Algorithm (QBGA) control, Shuand#64260;ed Frog Leaping Algorithm (SFLA) control, Cuckoo Search Algorithm (CSA) control, and the proposed Firefly Algorithm (FFA) control, so as to vary the duty cycle of the boost converter to enhance the eand#64256;ective operation of the system. The drawbacks associated with the existing optimization techniques such as QGBA, SFLA and CSA controls were they lag the search either in local optima and possess slow convergence rate. Hence to overcome the drawbacks associated with these methods, FFA optimization is developed, which works on the basis of the flashing light characteristics of the fireflies. The research modeling of this technique is done in MATLAB/SIMULINK 2014 platfo