Identification of coherent groups for dynamic analysis using modern techniques

AbstractBecause contemporary power systems are so large, it is not possible to model the complete system in details. For this reason the power system dynamic equivalent is established.Therefore, this thesis presents a new, simple and an accurate method to identify the coherent groups of generators and constructing the reduced order dynamic equivalent of power system.The proposed dynamic equivalent of power system is based on (1) Identification of coherent groups of generators, (2) Dynamic aggregation of generators in each of the coherent group and (3) Construct the reduced form of the transmission line network.The proposed method of identification of coherent generators is based on two different techniques; the first technique is based on six proposed coherency criterions introduced by using the time response of linearized power system model (ISEδ, ISEω, IAEδ, IAEω, MAEδ and MAEω). The second technique of the identification of coherent groups of generators is based on the application of fuzzy c-means clustering algorithm (FCM) as the coherency measures are derived from the time-domain response of generators to reveal the relations between any pair of generators then; finally; they are used as the initial membership matrix in the FCM clustering.The dynamic aggregation is obtained by replacing a group of coherent generators by an equivalent generator. The proposed form to calculate the parameters of the equivalent machine is based on the weighted mean values of the parameters of the machines included in the coherent group.The proposed form to calculate the new transmission lines paramete