A Crow Search Algorithm Approach for Optimal Reactive Power Dispatch with Multi-Type FACTS Devices

CSA (Crow Search Algorithm) based optimization technique has been used to find out the optimal setting of the reactive power sources for the loss minimum configuration of a connected power network. IEEE 57 bus test system was used as a standard power network and the purpose of the CSA was to determine the optimal size of the internal VAr sources i.e., the VAr sources already present in the network itself along with the proper placement of the FACTS (Flexible Alternating Current Transmission System) devices at the location already predefined. Here, three different kinds of FACTS (Flexible Alternating Current Transmission System) devices, SVC (Static Var Compensator), a shunt type, TCSC (Thyristor Controlled Series Compensator), a series type, UPFC (Unified Power Flow Controller), a facts device made of series and shunt controller were used as additional VAr sources. UPFC and TCSC positions were determined using power flow analysis. Lines carrying high amounts of reactive power were chosen for the placement of thyristor-controlled switched capacitors, whereas UPFCs were connected to the buses adjoining lines carrying both high amounts of active and reactive power. SVC positions were determined by weak node analysis. Hence, the proposed work may be conceived as a problem of reactive power optimization, where transmission system loss and overall system cost were minimized by the optimal coordination of existing VAr sources i.e. Q generations of generators, online tap changer transformer tap settings with the VArs injected into the system by different types of FACTS devices. Also, the Particle Swarm Optimization (PSO) method was run and the results were recorded. The proposed CSA (Crow Search Algorithm) method was compared with PSO (Particle Swarm optimization), HSA (Harmony Search Algorithm), and IHSA (Improved Harmony Search Algorithm) based techniques, and the superiority of the CSA-based method was established on the basis of the results.