Robust damping controller design for large power systems using wide area measurements

Conventional Power System Stabilizers (PSSs) are major local damping controllers acting through generator excitation systems. These conventional distributed local PSSs (LPSSs) are usually designed to have fixed parameters derived from a linearized model around a certain operating point. Final settings are made using field tests at several operating points. The quality of the system response is determined by many factors. One factor relates to the fact that the inherent non-linearity in the power system is a major source of model uncertainty. A second factor is that the system response depends on the selection of controller inputs that are utilized. It is found that if remote signals from one or more distant locations of the power system are applied to local controller design, system dynamic performance is enhanced. In this research, a supervisory level Power System Stabilizer (SPSS) using wide area measurements is proposed. It is possible to use H_∞ and LQG controllers with selected wide area measurements embedded into the SPSS control loop. This design accommodates power system nonlinear dynamic performance and model uncertainties. The control signals are sent to local machines to damp system oscillations through the machines' excitation systems. The robustness of the proposed controller is capable of compensating for the model uncertainty due to the nonlinear dynamic operation of the system and uncertain disturbances. The coordination of the robust SPSSs and local PSSs is implemented based on the principles of multi-agent system theory. Multi-agent theory is an active branch of applications in distributed artificial intelligence (DAI). Different to both the centralized and the decentralized control schemes, the multi-agent system employs a number of semi-autonomous agents to collaborate with each other to achieve a given task. The local PSSs have already been installed in the system and are tuned for damping local oscillation modes. Local PSSs are represented as local agents, which are placed at the selected generator excitation loops and remain in the same location throughout their working lives. The SPSS operates as a software agent. The agent contains a fuzzy logic controller switch to select the appropriate robust controller for the corresponding system operating condition. The performance of the robust controller as a power system stability agent is studied using a 29-machine 179-bus power system example. Based on testing under a linearized system environment, the simulation results show that the proposed robust controller can effectively damp system inter-area oscillations under a wide range of operating conditions. The implication of the wide area robust controller design is that higher level of power can be exchanged between areas using SPSS technology as compared to conventional PSS configurations.