Estimation of power system damping factors using pattern search

Interconnected power systems typically encounter low frequency oscillations during their operations. These oscillations are characterized by groups of coherent generators swinging against each other. When such undesired phenomenon is presented in a power system, it limits the amount of power transfer on the tie-lines among different regions containing the groups of coherent generators. This paper presents a new technique for analyzing the poorly damped low frequency oscillations in power system dynamic swings. The proposed algorithm is based on a meta-heuristic technique that has been getting added attention, i.e., Pattern Search Optimization. The method employs a digital set of measurements for power system dynamic swings to perform the analysis process digitally. The goal is to estimate the amount of damping in the swing curve as well as the oscillation frequency. The problem is formulated and presented as a dynamic estimation problem. The proposed technique is used to perform the estimation process by minimizing the total error associated with the estimated parameters. The algorithm is tested using different cases, including practical data for an existing power system. Results are evaluated and compared with other competing methods to show the potential of the proposed method.