Controller Tuning Using Genetic Algorithms

The increasing complexity of the modern control systems has emphasized the idea of applying new approaches in order to solve design problems for different control engineering problems. The genetic algorithms (GA) used in the paper are a promising heuristic approach to locating near-optimal tuning solutions. They are easy to implement and robust. This paper attempts to show how GAs can be applied for a two-step design problem. In the first step the GA assist the design of an ordinary PI controller for a plant with time delay, whereas in the second step GA help to tune the compensating controller for a structure with reference model. The presented cases study confirms that good performance can be achieved by the proposed method.