A neural-network-based controller for the cost-effective operation of a hybrid compensator for nonactive power

The need to eliminate distortion from power networks has led to the development of various compensator topologies. The increasing cost of electrical energy requires the choice of the most cost-effective compensator operation. An investigation of a neural-network-based controller that chooses the most cost-effective compensator mode of operation on the basis of a continuous analysis of load conditions and the operational losses of the elements in the compensator structure are reported. The modeling of operational losses of each subtopology and the required control strategy are discussed. The results show that the operational loss savings due to the neural-network-controlled hybrid compensator were 30%-70% as compared to the conventionally controlled hybrid compensator, while also conforming to other control strategy requirements.