A new method for optimization of Load Frequency Control parameters in multi-area power systems using Genetic Algorithms

Abstract

The main goal of the Load Frequency Control (LFC) is to establish a balance between load change (and losses) and supplied power in electrical power systems. The model of LFC used in the simulation must closely mimic real-life elements with minimal mistakes in terms of restrictions and generator behavior. To assess the model in the most accurate way, physical constraints such as Governor Dead Band, Generation Rate Constraint, and the Time delay in communication links should be imposed to demonstrate the effect of such constraints on the AGC performance. Introducing physical restraints and increasing the number of areas to be controlled results in increasing complexity. For such problems, data-driven intelligent methods like Genetic Algorithm (GA), Fuzzy Logic Controller, and Neural Networks have been used to find the optimal solution. GA has been used to tune the controller gains in several scientific papers concerning LFC and it is chosen as the optimization technique in this paper. The obtained results showed that using GA as an additional tool in LFC parameter optimization is a valuable method for stabilizing the LFC performance.