Oscillation damping improvement using PMU based PSS in power system

Abstract

The interconnection of power systems and the increase of power transfer are required to meet the increasing demand of electrical energy. These requirements drive the modern power systems closer to their limits of operation. Consequently, low frequency oscillations may arise in systems not affected previously and become a major source of systems instability and limitation factor of power exchange over long distances. One method of mitigation of those oscillations relies on power system stabilizers (PSSs) that can provide effective damping to local oscillation modes by using locally measured signals. However, they lack observability of inter-area modes, which results of poor damping for those modes. In this thesis, a wide area power system stabilizers (WAPSSs), able to effectively damp out inter-area oscillations in large power systems, is adopted. The proposed approach relies on the development of phasor measurement units (PMUs), which provide remote signals as inputs for WAPSSs for real-time implementation. The effectiveness of the proposed approach depends on many considerations. The first step is the selection of suitable control locations and remote signals with high observability to the targeted inter-area modes. To face this requirement, geometric measures of observability and controllability is adopted. The second crucial step is the optimal tuning of WAPSS to achieve adequate damping. A novel metaheuristic algorithm, comprehensive learning bat algorithm (CLBAT), is employed for the tuning process and compared with state of the art metaheuristic algorithms. Finally, the time delays associated with the global signals can deteriorate the performance of the proposed WAPSSs. Thereby a time delay compensators are proposed compensate time delays in wide area measurements based control. The performance of these WAPSSs has been evaluated on the New England/ New York and Two area four machine multimachine systems via eigenvalue analysis, and nonlinear time domain simulations