Load-Frequency Control: a GA based Bayesian Networks Multi-agent System

Bayesian Networks (BN) provides a robust probabilistic method of reasoningunder uncertainty. They have been successfully applied in a variety of real-world tasks butthey have received little attention in the area of load-frequency control (LFC). In practice,LFC systems use proportional-integral controllers. However since these controllers aredesigned using a linear model, the nonlinearities of the system are not accounted for andthey are incapable to gain good dynamical performance for a wide range of operatingconditions in a multi-area power system. A strategy for solving this problem due to thedistributed nature of a multi-area power system, is presented by using a BN multi-agentsystem. This method admits considerable flexibility in defining the control objective. AlsoBN provides a flexible means of representing and reasoning with probabilistic information.Efficient probabilistic inference algorithms in BN permit answering various probabilisticqueries about the system. Moreover using multi-agent structure in the proposed model,realized parallel computation and leading to a high degree of scalability. To demonstratethe capability of the proposed control structure, we construct a BN on the basis ofoptimized data using genetic algorithm (GA) for LFC of a three-area power system withtwo scenarios.